[GODE]final GODE code

Import

import numpy as np
import matplotlib.pyplot as plt
import matplotlib
import pandas as pd
import random

import warnings
warnings.simplefilter("ignore", np.ComplexWarning)
from haversine import haversine
from IPython.display import HTML
import plotly.graph_objects as go
import copy 

import tqdm
from rpy2.robjects.packages import importr
from rpy2.robjects.vectors import FloatVector 

from pygsp import graphs, filters, plotting, utils

from sklearn.metrics import confusion_matrix, precision_score, recall_score, f1_score, accuracy_score, roc_curve, auc

from pyod.models.lof import LOF
from pyod.models.knn import KNN
from pyod.models.cblof import CBLOF
from pyod.models.ocsvm import OCSVM
from pyod.models.mcd import MCD
from pyod.models.feature_bagging import FeatureBagging
from pyod.models.abod import ABOD
from pyod.models.iforest import IForest
from pyod.models.hbos import HBOS
from pyod.models.sos import SOS
from pyod.models.so_gaal import SO_GAAL
from pyod.models.mo_gaal import MO_GAAL
from pyod.models.lscp import LSCP

2023-11-27 13:23:42.982919: I tensorflow/core/util/port.cc:110] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.

Class

class Conf_matrx:
    def __init__(self,original,compare):
        self.original = original
        self.compare = compare
    def conf(self,name):
        self.name = name
        self.conf_matrix = confusion_matrix(self.original, self.compare)
        
        fig, ax = plt.subplots(figsize=(5, 5))
        ax.matshow(self.conf_matrix, cmap=plt.cm.Oranges, alpha=0.3)
        for i in range(self.conf_matrix.shape[0]):
            for j in range(self.conf_matrix.shape[1]):
                ax.text(x=j, y=i,s=self.conf_matrix[i, j], va='center', ha='center', size='xx-large')
        plt.xlabel('Predictions', fontsize=18)
        plt.ylabel('Actuals', fontsize=18)
        plt.title('Confusion Matrix of ' + str(name), fontsize=18)
        plt.show()
        
        self.acc = accuracy_score(self.original, self.compare)
        self.pre = precision_score(self.original, self.compare)
        self.rec = recall_score(self.original, self.compare)
        self.f1 = f1_score(self.original, self.compare)
        
        print('Accuracy: %.3f' % self.acc)
        print('Precision: %.3f' % self.pre)
        print('Recall: %.3f' % self.rec)
        print('F1 Score: %.3f' % self.f1)

class Linear:
    def __init__(self,df):
        self.df = df
        self.y = df.y.to_numpy()
        self.x = df.x.to_numpy()
        self.n = len(self.y)
        self.W = w
    def _eigen(self):
        d= self.W.sum(axis=1)
        D= np.diag(d)
        self.L = np.diag(1/np.sqrt(d)) @ (D-self.W) @ np.diag(1/np.sqrt(d))
        self.lamb, self.Psi = np.linalg.eigh(self.L)
        self.Lamb = np.diag(self.lamb)      
    def fit(self,sd=20): # fit with ebayesthresh
        self._eigen()
        self.ybar = self.Psi.T @ self.y # fbar := graph fourier transform of f
        self.power = self.ybar**2 
        ebayesthresh = importr('EbayesThresh').ebayesthresh
        self.power_threshed=np.array(ebayesthresh(FloatVector(self.power),sd=sd))
        self.ybar_threshed = np.where(self.power_threshed>0,self.ybar,0)
        self.yhat = self.Psi@self.ybar_threshed
        self.df = self.df.assign(yHat = self.yhat)
        self.df = self.df.assign(Residual = self.df.y- self.df.yHat)

class Orbit:
    def __init__(self,df):
        self.df = df 
        self.f = df.f.to_numpy()
        self.x = df.x.to_numpy()
        self.y = df.y.to_numpy()
        self.n = len(self.f)
        self.theta= None
    def get_distance(self):
        self.D = np.zeros([self.n,self.n])
        locations = np.stack([self.x, self.y],axis=1)
        for i in tqdm.tqdm(range(self.n)):
            for j in range(i,self.n):
                self.D[i,j]=np.linalg.norm(locations[i]-locations[j])
        self.D = self.D + self.D.T
    def get_weightmatrix(self,theta=1,beta=0.5,kappa=4000):
        self.theta = theta
        dist = np.where(self.D < kappa,self.D,0)
        self.W = np.exp(-(dist/self.theta)**2)
    def _eigen(self):
        d= self.W.sum(axis=1)
        D= np.diag(d)
        self.L = np.diag(1/np.sqrt(d)) @ (D-self.W) @ np.diag(1/np.sqrt(d))
        self.lamb, self.Psi = np.linalg.eigh(self.L)
        self.Lamb = np.diag(self.lamb)       
    def fit(self,sd=5): # fit with ebayesthresh
        self._eigen()
        self.fbar = self.Psi.T @ self.f # fbar := graph fourier transform of f
        self.power = self.fbar**2 
        ebayesthresh = importr('EbayesThresh').ebayesthresh
        self.power_threshed=np.array(ebayesthresh(FloatVector(self.power),sd=sd))
        self.fbar_threshed = np.where(self.power_threshed>0,self.fbar,0)
        self.fhat = self.Psi@self.fbar_threshed
        self.df = self.df.assign(fHat = self.fhat)
        self.df = self.df.assign(Residual = self.df.f- self.df.fHat)

class BUNNY:
    def __init__(self,df):
        self.df = df 
        self.f = df.f.to_numpy()
        self.z = df.z.to_numpy()
        self.x = df.x.to_numpy()
        self.y = df.y.to_numpy()
        self.noise = df.noise.to_numpy()
        self.fnoise = self.f + self.noise
        self.W = _W
        self.n = len(self.f)
        self.theta= None
    def _eigen(self):
        d= self.W.sum(axis=1)
        D= np.diag(d)
        self.L = np.diag(1/np.sqrt(d)) @ (D-self.W) @ np.diag(1/np.sqrt(d))
        self.lamb, self.Psi = np.linalg.eigh(self.L)
        self.Lamb = np.diag(self.lamb)       
    def fit(self,sd=5): # fit with ebayesthresh
        self._eigen()
        self.fbar = self.Psi.T @ self.fnoise # fbar := graph fourier transform of f
        self.power = self.fbar**2 
        ebayesthresh = importr('EbayesThresh').ebayesthresh
        self.power_threshed=np.array(ebayesthresh(FloatVector(self.power),sd=sd))
        self.fbar_threshed = np.where(self.power_threshed>0,self.fbar,0)
        self.fhat = self.Psi@self.fbar_threshed
        self.df = self.df.assign(fnoise = self.fnoise)
        self.df = self.df.assign(fHat = self.fhat)
        self.df = self.df.assign(Residual = self.df.f + self.df.noise - self.df.fHat)

tab_linear = pd.DataFrame(columns=["Accuracy","Precision","Recall","F1","AUC"])
tab_orbit = pd.DataFrame(columns=["Accuracy","Precision","Recall","F1","AUC"])
tab_bunny = pd.DataFrame(columns=["Accuracy","Precision","Recall","F1","AUC"])

Linear

np.random.seed(6)
epsilon = np.around(np.random.normal(size=1000),15)
signal = np.random.choice(np.concatenate((np.random.uniform(-7, -5, 25).round(15), np.random.uniform(5, 7, 25).round(15), np.repeat(0, 950))), 1000)
eta = signal + epsilon
outlier_true_linear= signal.copy()
outlier_true_linear = list(map(lambda x: 1 if x!=0 else 0,outlier_true_linear))

x_1 = np.linspace(0,2,1000)
y1_1 = 5 * x_1
y_1 = y1_1 + eta # eta = signal + epsilon
_df=pd.DataFrame({'x':x_1, 'y':y_1})

w=np.zeros((1000,1000))
for i in range(1000):
    for j in range(1000):
        if i==j :
            w[i,j] = 0
        elif np.abs(i-j) <= 1 : 
            w[i,j] = 1

GODE_Linear

_Linear = Linear(_df)
_Linear.fit(sd=20)

outlier_GODE_linear_old = (_Linear.df['Residual']**2).tolist()
sorted_data = sorted(outlier_GODE_linear_old,reverse=True)
index = int(len(sorted_data) * 0.05)
five_percent = sorted_data[index]
outlier_GODE_linear = list(map(lambda x: 1 if x > five_percent else 0,outlier_GODE_linear_old))

_conf = Conf_matrx(outlier_true_linear,outlier_GODE_linear)
_conf.conf('GODE')

Accuracy: 0.999
Precision: 1.000
Recall: 0.980
F1 Score: 0.990

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.999
Precision(TP / TP + FP): 1.000
Recall(TP / TP + FN): 0.980
F1 Score(2*precision*recall/precision+recall): 0.990

fpr, tpr, thresh = roc_curve(outlier_true_linear,outlier_GODE_linear_old)

auc(fpr, tpr)

0.999979338416083

tab_linear = pd.concat([tab_linear,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_linear

	Accuracy	Precision	Recall	F1	AUC
GODE	0.999	1.0	0.980392	0.990099	0.999979

LOF_Linear

np.random.seed(77)
clf = LOF(contamination=0.05)
clf.fit(_df[['x', 'y']])

LOF(algorithm='auto', contamination=0.05, leaf_size=30, metric='minkowski',
  metric_params=None, n_jobs=1, n_neighbors=20, novelty=True, p=2)

outlier_LOF_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_linear,clf.fit_predict(_df))
_conf.conf("LOF (Breunig et al., 2000)")

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/utils/deprecation.py:86: FutureWarning: Function fit_predict is deprecated
  warnings.warn(msg, category=FutureWarning)

Accuracy: 0.991
Precision: 0.920
Recall: 0.902
F1 Score: 0.911

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.991
Precision(TP / TP + FP): 0.920
Recall(TP / TP + FN): 0.902
F1 Score(2*precision*recall/precision+recall): 0.911

fpr, tpr, thresh = roc_curve(outlier_true_linear,clf.decision_function(_df))

auc(fpr, tpr)

0.9975412715138742

tab_linear = pd.concat([tab_linear,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_linear

	Accuracy	Precision	Recall	F1	AUC
GODE	0.999	1.00	0.980392	0.990099	0.999979
LOF (Breunig et al., 2000)	0.991	0.92	0.901961	0.910891	0.997541

KNN_Linear

np.random.seed(77)
clf = KNN(contamination=0.05)
clf.fit(_df[['x', 'y']])

KNN(algorithm='auto', contamination=0.05, leaf_size=30, method='largest',
  metric='minkowski', metric_params=None, n_jobs=1, n_neighbors=5, p=2,
  radius=1.0)

outlier_KNN_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_linear,outlier_KNN_one)
_conf.conf("kNN (Ramaswamy et al., 2000)")

Accuracy: 0.991
Precision: 0.920
Recall: 0.902
F1 Score: 0.911

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.991
Precision(TP / TP + FP): 0.920
Recall(TP / TP + FN): 0.902
F1 Score(2*precision*recall/precision+recall): 0.911

fpr, tpr, thresh = roc_curve(outlier_true_linear,clf.decision_function(_df))

auc(fpr, tpr)

0.9973656480505795

tab_linear = pd.concat([tab_linear,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_linear

	Accuracy	Precision	Recall	F1	AUC
GODE	0.999	1.00	0.980392	0.990099	0.999979
LOF (Breunig et al., 2000)	0.991	0.92	0.901961	0.910891	0.997541
kNN (Ramaswamy et al., 2000)	0.991	0.92	0.901961	0.910891	0.997366

CBLOF_Linear

clf = CBLOF(contamination=0.05,random_state=77)
clf.fit(_df[['x', 'y']])

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning
  super()._check_params_vs_input(X, default_n_init=10)

CBLOF(alpha=0.9, beta=5, check_estimator=False, clustering_estimator=None,
   contamination=0.05, n_clusters=8, n_jobs=None, random_state=77,
   use_weights=False)

outlier_CBLOF_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_linear,outlier_CBLOF_one)
_conf.conf("CBLOF (He et al., 2003)")

Accuracy: 0.969
Precision: 0.700
Recall: 0.686
F1 Score: 0.693

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.969
Precision(TP / TP + FP): 0.700
Recall(TP / TP + FN): 0.686
F1 Score(2*precision*recall/precision+recall): 0.693

fpr, tpr, thresh = roc_curve(outlier_true_linear,clf.decision_function(_df))

auc(fpr, tpr)

0.9592140333477964

tab_linear = pd.concat([tab_linear,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_linear

	Accuracy	Precision	Recall	F1	AUC
GODE	0.999	1.00	0.980392	0.990099	0.999979
LOF (Breunig et al., 2000)	0.991	0.92	0.901961	0.910891	0.997541
kNN (Ramaswamy et al., 2000)	0.991	0.92	0.901961	0.910891	0.997366
CBLOF (He et al., 2003)	0.969	0.70	0.686275	0.693069	0.959214

OCSVM_Linear

np.random.seed(77)
clf = OCSVM(nu=0.05)
clf.fit(_df)

OCSVM(cache_size=200, coef0=0.0, contamination=0.1, degree=3, gamma='auto',
   kernel='rbf', max_iter=-1, nu=0.05, shrinking=True, tol=0.001,
   verbose=False)

outlier_OSVM_one = list(clf.predict(_df))

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/base.py:457: UserWarning: X has feature names, but OneClassSVM was fitted without feature names
  warnings.warn(

_conf = Conf_matrx(outlier_true_linear,outlier_OSVM_one)
_conf.conf("OCSVM (Sch ̈olkopf et al., 2001)")

Accuracy: 0.923
Precision: 0.370
Recall: 0.725
F1 Score: 0.490

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.923
Precision(TP / TP + FP): 0.370
Recall(TP / TP + FN): 0.725
F1 Score(2*precision*recall/precision+recall): 0.490

fpr, tpr, thresh = roc_curve(outlier_true_linear,clf.decision_function(_df))

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/base.py:457: UserWarning: X has feature names, but OneClassSVM was fitted without feature names
  warnings.warn(

auc(fpr, tpr)

0.8641500857455733

tab_linear = pd.concat([tab_linear,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_linear

	Accuracy	Precision	Recall	F1	AUC
GODE	0.999	1.00	0.980392	0.990099	0.999979
LOF (Breunig et al., 2000)	0.991	0.92	0.901961	0.910891	0.997541
kNN (Ramaswamy et al., 2000)	0.991	0.92	0.901961	0.910891	0.997366
CBLOF (He et al., 2003)	0.969	0.70	0.686275	0.693069	0.959214
OCSVM (Sch ̈olkopf et al., 2001)	0.923	0.37	0.725490	0.490066	0.864150

MCD_Linear

clf = MCD(contamination=0.05, random_state = 77)
clf.fit(_df[['x', 'y']])

MCD(assume_centered=False, contamination=0.05, random_state=77,
  store_precision=True, support_fraction=None)

outlier_MCD_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_linear,outlier_MCD_one)
_conf.conf("MCD (Hardin and Rocke, 2004)")

Accuracy: 0.999
Precision: 1.000
Recall: 0.980
F1 Score: 0.990

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.999
Precision(TP / TP + FP): 1.000
Recall(TP / TP + FN): 0.980
F1 Score(2*precision*recall/precision+recall): 0.990

fpr, tpr, thresh = roc_curve(outlier_true_linear,clf.decision_function(_df))

auc(fpr, tpr)

0.999958676832166

tab_linear = pd.concat([tab_linear,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_linear

	Accuracy	Precision	Recall	F1	AUC
GODE	0.999	1.00	0.980392	0.990099	0.999979
LOF (Breunig et al., 2000)	0.991	0.92	0.901961	0.910891	0.997541
kNN (Ramaswamy et al., 2000)	0.991	0.92	0.901961	0.910891	0.997366
CBLOF (He et al., 2003)	0.969	0.70	0.686275	0.693069	0.959214
OCSVM (Sch ̈olkopf et al., 2001)	0.923	0.37	0.725490	0.490066	0.864150
MCD (Hardin and Rocke, 2004)	0.999	1.00	0.980392	0.990099	0.999959

Feature Bagging_Linear

clf = FeatureBagging(contamination=0.05, random_state=77)
clf.fit(_df[['x', 'y']])

FeatureBagging(base_estimator=None, bootstrap_features=False,
        check_detector=True, check_estimator=False, combination='average',
        contamination=0.05, estimator_params={}, max_features=1.0,
        n_estimators=10, n_jobs=1, random_state=77, verbose=0)

outlier_FeatureBagging_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_linear,outlier_FeatureBagging_one)
_conf.conf("Feature Bagging (Lazarevic and Kumar, 2005)")

Accuracy: 0.993
Precision: 0.940
Recall: 0.922
F1 Score: 0.931

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.993
Precision(TP / TP + FP): 0.940
Recall(TP / TP + FN): 0.922
F1 Score(2*precision*recall/precision+recall): 0.931

fpr, tpr, thresh = roc_curve(outlier_true_linear,clf.decision_function(_df))

auc(fpr, tpr)

0.9973966404264551

tab_linear = pd.concat([tab_linear,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_linear

	Accuracy	Precision	Recall	F1	AUC
GODE	0.999	1.00	0.980392	0.990099	0.999979
LOF (Breunig et al., 2000)	0.991	0.92	0.901961	0.910891	0.997541
kNN (Ramaswamy et al., 2000)	0.991	0.92	0.901961	0.910891	0.997366
CBLOF (He et al., 2003)	0.969	0.70	0.686275	0.693069	0.959214
OCSVM (Sch ̈olkopf et al., 2001)	0.923	0.37	0.725490	0.490066	0.864150
MCD (Hardin and Rocke, 2004)	0.999	1.00	0.980392	0.990099	0.999959
Feature Bagging (Lazarevic and Kumar, 2005)	0.993	0.94	0.921569	0.930693	0.997397

ABOD_Linear

np.random.seed(77)
clf = ABOD(contamination=0.05)
clf.fit(_df[['x', 'y']])

ABOD(contamination=0.05, method='fast', n_neighbors=5)

outlier_ABOD_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_linear,outlier_ABOD_one)
_conf.conf("ABOD (Kriegel et al., 2008)")

Accuracy: 0.973
Precision: 0.740
Recall: 0.725
F1 Score: 0.733

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.973
Precision(TP / TP + FP): 0.740
Recall(TP / TP + FN): 0.725
F1 Score(2*precision*recall/precision+recall): 0.733

fpr, tpr, thresh = roc_curve(outlier_true_linear,clf.decision_function(_df))

auc(fpr, tpr)

0.9902064092233311

tab_linear = pd.concat([tab_linear,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_linear

	Accuracy	Precision	Recall	F1	AUC
GODE	0.999	1.00	0.980392	0.990099	0.999979
LOF (Breunig et al., 2000)	0.991	0.92	0.901961	0.910891	0.997541
kNN (Ramaswamy et al., 2000)	0.991	0.92	0.901961	0.910891	0.997366
CBLOF (He et al., 2003)	0.969	0.70	0.686275	0.693069	0.959214
OCSVM (Sch ̈olkopf et al., 2001)	0.923	0.37	0.725490	0.490066	0.864150
MCD (Hardin and Rocke, 2004)	0.999	1.00	0.980392	0.990099	0.999959
Feature Bagging (Lazarevic and Kumar, 2005)	0.993	0.94	0.921569	0.930693	0.997397
ABOD (Kriegel et al., 2008)	0.973	0.74	0.725490	0.732673	0.990206

IForest_Linear

clf = IForest(contamination=0.05,random_state=77)
clf.fit(_df[['x', 'y']])

IForest(behaviour='old', bootstrap=False, contamination=0.05,
    max_features=1.0, max_samples='auto', n_estimators=100, n_jobs=1,
    random_state=77, verbose=0)

outlier_IForest_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_linear,outlier_IForest_one)
_conf.conf("Isolation Forest (Liu et al., 2008)")

Accuracy: 0.987
Precision: 0.880
Recall: 0.863
F1 Score: 0.871

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.987
Precision(TP / TP + FP): 0.880
Recall(TP / TP + FN): 0.863
F1 Score(2*precision*recall/precision+recall): 0.871

fpr, tpr, thresh = roc_curve(outlier_true_linear,clf.decision_function(_df))

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/base.py:457: UserWarning: X has feature names, but IsolationForest was fitted without feature names
  warnings.warn(

auc(fpr, tpr)

0.9958263600487613

tab_linear = pd.concat([tab_linear,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_linear

	Accuracy	Precision	Recall	F1	AUC
GODE	0.999	1.00	0.980392	0.990099	0.999979
LOF (Breunig et al., 2000)	0.991	0.92	0.901961	0.910891	0.997541
kNN (Ramaswamy et al., 2000)	0.991	0.92	0.901961	0.910891	0.997366
CBLOF (He et al., 2003)	0.969	0.70	0.686275	0.693069	0.959214
OCSVM (Sch ̈olkopf et al., 2001)	0.923	0.37	0.725490	0.490066	0.864150
MCD (Hardin and Rocke, 2004)	0.999	1.00	0.980392	0.990099	0.999959
Feature Bagging (Lazarevic and Kumar, 2005)	0.993	0.94	0.921569	0.930693	0.997397
ABOD (Kriegel et al., 2008)	0.973	0.74	0.725490	0.732673	0.990206
Isolation Forest (Liu et al., 2008)	0.987	0.88	0.862745	0.871287	0.995826

HBOS_Linear

np.random.seed(77)
clf = HBOS(contamination=0.05)
clf.fit(_df[['x', 'y']])

HBOS(alpha=0.1, contamination=0.05, n_bins=10, tol=0.5)

outlier_HBOS_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_linear,outlier_HBOS_one)
_conf.conf("HBOS (Goldstein and Dengel, 2012)")

Accuracy: 0.972
Precision: 0.926
Recall: 0.490
F1 Score: 0.641

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.972
Precision(TP / TP + FP): 0.926
Recall(TP / TP + FN): 0.490
F1 Score(2*precision*recall/precision+recall): 0.641

fpr, tpr, thresh = roc_curve(outlier_true_linear,clf.decision_function(_df))

auc(fpr, tpr)

0.8636438769396062

tab_linear = pd.concat([tab_linear,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_linear

	Accuracy	Precision	Recall	F1	AUC
GODE	0.999	1.000000	0.980392	0.990099	0.999979
LOF (Breunig et al., 2000)	0.991	0.920000	0.901961	0.910891	0.997541
kNN (Ramaswamy et al., 2000)	0.991	0.920000	0.901961	0.910891	0.997366
CBLOF (He et al., 2003)	0.969	0.700000	0.686275	0.693069	0.959214
OCSVM (Sch ̈olkopf et al., 2001)	0.923	0.370000	0.725490	0.490066	0.864150
MCD (Hardin and Rocke, 2004)	0.999	1.000000	0.980392	0.990099	0.999959
Feature Bagging (Lazarevic and Kumar, 2005)	0.993	0.940000	0.921569	0.930693	0.997397
ABOD (Kriegel et al., 2008)	0.973	0.740000	0.725490	0.732673	0.990206
Isolation Forest (Liu et al., 2008)	0.987	0.880000	0.862745	0.871287	0.995826
HBOS (Goldstein and Dengel, 2012)	0.972	0.925926	0.490196	0.641026	0.863644

SOS_Linear

np.random.seed(77)
clf = SOS(contamination=0.05)
clf.fit(_df[['x', 'y']])

SOS(contamination=0.05, eps=1e-05, metric='euclidean', perplexity=4.5)

outlier_SOS_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_linear,outlier_SOS_one)
_conf.conf("SOS (Janssens et al., 2012)")

Accuracy: 0.907
Precision: 0.080
Recall: 0.078
F1 Score: 0.079

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.907
Precision(TP / TP + FP): 0.080
Recall(TP / TP + FN): 0.078
F1 Score(2*precision*recall/precision+recall): 0.079

fpr, tpr, thresh = roc_curve(outlier_true_linear,clf.decision_function(_df))

auc(fpr, tpr)

0.5417054071365112

tab_linear = pd.concat([tab_linear,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_linear

	Accuracy	Precision	Recall	F1	AUC
GODE	0.999	1.000000	0.980392	0.990099	0.999979
LOF (Breunig et al., 2000)	0.991	0.920000	0.901961	0.910891	0.997541
kNN (Ramaswamy et al., 2000)	0.991	0.920000	0.901961	0.910891	0.997366
CBLOF (He et al., 2003)	0.969	0.700000	0.686275	0.693069	0.959214
OCSVM (Sch ̈olkopf et al., 2001)	0.923	0.370000	0.725490	0.490066	0.864150
MCD (Hardin and Rocke, 2004)	0.999	1.000000	0.980392	0.990099	0.999959
Feature Bagging (Lazarevic and Kumar, 2005)	0.993	0.940000	0.921569	0.930693	0.997397
ABOD (Kriegel et al., 2008)	0.973	0.740000	0.725490	0.732673	0.990206
Isolation Forest (Liu et al., 2008)	0.987	0.880000	0.862745	0.871287	0.995826
HBOS (Goldstein and Dengel, 2012)	0.972	0.925926	0.490196	0.641026	0.863644
SOS (Janssens et al., 2012)	0.907	0.080000	0.078431	0.079208	0.541705

SO_GAAL_Linear

np.random.seed(77)
clf = SO_GAAL(contamination=0.05)
clf.fit(_df[['x', 'y']])

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/keras/src/optimizers/legacy/gradient_descent.py:114: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.
  super().__init__(name, **kwargs)

Epoch 1 of 60

Testing for epoch 1 index 1:

Testing for epoch 1 index 2:
Epoch 2 of 60

Testing for epoch 2 index 1:

Testing for epoch 2 index 2:
Epoch 3 of 60

Testing for epoch 3 index 1:

Testing for epoch 3 index 2:
Epoch 4 of 60

Testing for epoch 4 index 1:

Testing for epoch 4 index 2:
Epoch 5 of 60

Testing for epoch 5 index 1:

Testing for epoch 5 index 2:
Epoch 6 of 60

Testing for epoch 6 index 1:

Testing for epoch 6 index 2:
Epoch 7 of 60

Testing for epoch 7 index 1:

Testing for epoch 7 index 2:
Epoch 8 of 60

Testing for epoch 8 index 1:

Testing for epoch 8 index 2:
Epoch 9 of 60

Testing for epoch 9 index 1:

Testing for epoch 9 index 2:
Epoch 10 of 60

Testing for epoch 10 index 1:

Testing for epoch 10 index 2:
Epoch 11 of 60

Testing for epoch 11 index 1:

Testing for epoch 11 index 2:
Epoch 12 of 60

Testing for epoch 12 index 1:

Testing for epoch 12 index 2:
Epoch 13 of 60

Testing for epoch 13 index 1:

Testing for epoch 13 index 2:
Epoch 14 of 60

Testing for epoch 14 index 1:

Testing for epoch 14 index 2:
Epoch 15 of 60

Testing for epoch 15 index 1:

Testing for epoch 15 index 2:
Epoch 16 of 60

Testing for epoch 16 index 1:

Testing for epoch 16 index 2:
Epoch 17 of 60

Testing for epoch 17 index 1:

Testing for epoch 17 index 2:
Epoch 18 of 60

Testing for epoch 18 index 1:

Testing for epoch 18 index 2:
Epoch 19 of 60

Testing for epoch 19 index 1:

Testing for epoch 19 index 2:
Epoch 20 of 60

Testing for epoch 20 index 1:

Testing for epoch 20 index 2:
Epoch 21 of 60

Testing for epoch 21 index 1:

Testing for epoch 21 index 2:
Epoch 22 of 60

Testing for epoch 22 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.1165

Testing for epoch 22 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 1.1189
Epoch 23 of 60

Testing for epoch 23 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.1783

Testing for epoch 23 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 1.1937
Epoch 24 of 60

Testing for epoch 24 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.2216

Testing for epoch 24 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 1.2049
Epoch 25 of 60

Testing for epoch 25 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.2399

Testing for epoch 25 index 2:
16/16 [==============================] - 0s 909us/step - loss: 1.2524
Epoch 26 of 60

Testing for epoch 26 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.2894

Testing for epoch 26 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.3202
Epoch 27 of 60

Testing for epoch 27 index 1:
16/16 [==============================] - 0s 811us/step - loss: 1.2910

Testing for epoch 27 index 2:
16/16 [==============================] - 0s 815us/step - loss: 1.3322
Epoch 28 of 60

Testing for epoch 28 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.3269

Testing for epoch 28 index 2:
16/16 [==============================] - 0s 810us/step - loss: 1.3402
Epoch 29 of 60

Testing for epoch 29 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 1.3536

Testing for epoch 29 index 2:
16/16 [==============================] - 0s 789us/step - loss: 1.4119
Epoch 30 of 60

Testing for epoch 30 index 1:
16/16 [==============================] - 0s 809us/step - loss: 1.4078

Testing for epoch 30 index 2:
16/16 [==============================] - 0s 793us/step - loss: 1.3819
Epoch 31 of 60

Testing for epoch 31 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 1.4048

Testing for epoch 31 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 1.4149
Epoch 32 of 60

Testing for epoch 32 index 1:
16/16 [==============================] - 0s 774us/step - loss: 1.4285

Testing for epoch 32 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.4326
Epoch 33 of 60

Testing for epoch 33 index 1:
16/16 [==============================] - 0s 790us/step - loss: 1.4830

Testing for epoch 33 index 2:
16/16 [==============================] - 0s 844us/step - loss: 1.4463
Epoch 34 of 60

Testing for epoch 34 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.4626

Testing for epoch 34 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.4730
Epoch 35 of 60

Testing for epoch 35 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.4690

Testing for epoch 35 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.5170
Epoch 36 of 60

Testing for epoch 36 index 1:
16/16 [==============================] - 0s 841us/step - loss: 1.5280

Testing for epoch 36 index 2:
16/16 [==============================] - 0s 907us/step - loss: 1.5116
Epoch 37 of 60

Testing for epoch 37 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.5294

Testing for epoch 37 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.5397
Epoch 38 of 60

Testing for epoch 38 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.5235

Testing for epoch 38 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 1.5700
Epoch 39 of 60

Testing for epoch 39 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 1.5820

Testing for epoch 39 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.5393
Epoch 40 of 60

Testing for epoch 40 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.5693

Testing for epoch 40 index 2:
16/16 [==============================] - 0s 959us/step - loss: 1.5822
Epoch 41 of 60

Testing for epoch 41 index 1:
16/16 [==============================] - 0s 869us/step - loss: 1.6224

Testing for epoch 41 index 2:
16/16 [==============================] - 0s 964us/step - loss: 1.6510
Epoch 42 of 60

Testing for epoch 42 index 1:
16/16 [==============================] - 0s 855us/step - loss: 1.6076

Testing for epoch 42 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 1.5986
Epoch 43 of 60

Testing for epoch 43 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.6366

Testing for epoch 43 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 1.6675
Epoch 44 of 60

Testing for epoch 44 index 1:
16/16 [==============================] - 0s 839us/step - loss: 1.6276

Testing for epoch 44 index 2:
16/16 [==============================] - 0s 942us/step - loss: 1.6894
Epoch 45 of 60

Testing for epoch 45 index 1:
16/16 [==============================] - 0s 779us/step - loss: 1.7096

Testing for epoch 45 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.7009
Epoch 46 of 60

Testing for epoch 46 index 1:
16/16 [==============================] - 0s 950us/step - loss: 1.7469

Testing for epoch 46 index 2:
16/16 [==============================] - 0s 802us/step - loss: 1.7162
Epoch 47 of 60

Testing for epoch 47 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.6864

Testing for epoch 47 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.7342
Epoch 48 of 60

Testing for epoch 48 index 1:
16/16 [==============================] - 0s 850us/step - loss: 1.7271

Testing for epoch 48 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.8060
Epoch 49 of 60

Testing for epoch 49 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.7241

Testing for epoch 49 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.8112
Epoch 50 of 60

Testing for epoch 50 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 1.7575

Testing for epoch 50 index 2:
16/16 [==============================] - 0s 797us/step - loss: 1.7625
Epoch 51 of 60

Testing for epoch 51 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 1.7826

Testing for epoch 51 index 2:
16/16 [==============================] - 0s 852us/step - loss: 1.7773
Epoch 52 of 60

Testing for epoch 52 index 1:
16/16 [==============================] - 0s 845us/step - loss: 1.7978

Testing for epoch 52 index 2:
16/16 [==============================] - 0s 948us/step - loss: 1.8201
Epoch 53 of 60

Testing for epoch 53 index 1:
16/16 [==============================] - 0s 996us/step - loss: 1.8171

Testing for epoch 53 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.8348
Epoch 54 of 60

Testing for epoch 54 index 1:
16/16 [==============================] - 0s 788us/step - loss: 1.8149

Testing for epoch 54 index 2:
16/16 [==============================] - 0s 795us/step - loss: 1.8123
Epoch 55 of 60

Testing for epoch 55 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 1.8203

Testing for epoch 55 index 2:
16/16 [==============================] - 0s 759us/step - loss: 1.8700
Epoch 56 of 60

Testing for epoch 56 index 1:
16/16 [==============================] - 0s 822us/step - loss: 1.8608

Testing for epoch 56 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.8683
Epoch 57 of 60

Testing for epoch 57 index 1:
16/16 [==============================] - 0s 810us/step - loss: 1.9068

Testing for epoch 57 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.9222
Epoch 58 of 60

Testing for epoch 58 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.8711

Testing for epoch 58 index 2:
16/16 [==============================] - 0s 995us/step - loss: 1.8794
Epoch 59 of 60

Testing for epoch 59 index 1:
16/16 [==============================] - 0s 784us/step - loss: 1.9300

Testing for epoch 59 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.8748
Epoch 60 of 60

Testing for epoch 60 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.9501

Testing for epoch 60 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.9530
32/32 [==============================] - 0s 1ms/step

SO_GAAL(contamination=0.05, lr_d=0.01, lr_g=0.0001, momentum=0.9,
    stop_epochs=20)

outlier_SO_GAAL_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_linear,outlier_SO_GAAL_one)
_conf.conf("SO-GAAL (Liu et al., 2019)")

Accuracy: 0.946
Precision: 0.468
Recall: 0.431
F1 Score: 0.449

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.946
Precision(TP / TP + FP): 0.468
Recall(TP / TP + FN): 0.431
F1 Score(2*precision*recall/precision+recall): 0.449

fpr, tpr, thresh = roc_curve(outlier_true_linear,clf.decision_function(_df))

32/32 [==============================] - 0s 823us/step

auc(fpr, tpr)

0.575208165457964

tab_linear = pd.concat([tab_linear,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_linear

	Accuracy	Precision	Recall	F1	AUC
GODE	0.999	1.000000	0.980392	0.990099	0.999979
LOF (Breunig et al., 2000)	0.991	0.920000	0.901961	0.910891	0.997541
kNN (Ramaswamy et al., 2000)	0.991	0.920000	0.901961	0.910891	0.997366
CBLOF (He et al., 2003)	0.969	0.700000	0.686275	0.693069	0.959214
OCSVM (Sch ̈olkopf et al., 2001)	0.923	0.370000	0.725490	0.490066	0.864150
MCD (Hardin and Rocke, 2004)	0.999	1.000000	0.980392	0.990099	0.999959
Feature Bagging (Lazarevic and Kumar, 2005)	0.993	0.940000	0.921569	0.930693	0.997397
ABOD (Kriegel et al., 2008)	0.973	0.740000	0.725490	0.732673	0.990206
Isolation Forest (Liu et al., 2008)	0.987	0.880000	0.862745	0.871287	0.995826
HBOS (Goldstein and Dengel, 2012)	0.972	0.925926	0.490196	0.641026	0.863644
SOS (Janssens et al., 2012)	0.907	0.080000	0.078431	0.079208	0.541705
SO-GAAL (Liu et al., 2019)	0.946	0.468085	0.431373	0.448980	0.575208

MO_GAAL_Linear

np.random.seed(77)
clf = MO_GAAL(contamination=0.05)
clf.fit(_df[['x', 'y']])

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/keras/src/optimizers/legacy/gradient_descent.py:114: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.
  super().__init__(name, **kwargs)

Epoch 1 of 60

Testing for epoch 1 index 1:
WARNING:tensorflow:5 out of the last 74 calls to <function Model.make_predict_function.<locals>.predict_function at 0x7fd6d0525000> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has reduce_retracing=True option that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.
WARNING:tensorflow:5 out of the last 12 calls to <function Model.make_predict_function.<locals>.predict_function at 0x7fd7f68d8e50> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has reduce_retracing=True option that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.
32/32 [==============================] - 0s 1ms/step
WARNING:tensorflow:5 out of the last 165 calls to <function Model.make_train_function.<locals>.train_function at 0x7fd6d014c0d0> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has reduce_retracing=True option that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.
WARNING:tensorflow:6 out of the last 166 calls to <function Model.make_train_function.<locals>.train_function at 0x7fd7ee5e8c10> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has reduce_retracing=True option that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.

Testing for epoch 1 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 2 of 60

Testing for epoch 2 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 2 index 2:
32/32 [==============================] - 0s 799us/step
Epoch 3 of 60

Testing for epoch 3 index 1:
32/32 [==============================] - 0s 590us/step

Testing for epoch 3 index 2:
32/32 [==============================] - 0s 799us/step
Epoch 4 of 60

Testing for epoch 4 index 1:
32/32 [==============================] - 0s 576us/step

Testing for epoch 4 index 2:
32/32 [==============================] - 0s 574us/step
Epoch 5 of 60

Testing for epoch 5 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 5 index 2:
32/32 [==============================] - 0s 588us/step
Epoch 6 of 60

Testing for epoch 6 index 1:
32/32 [==============================] - 0s 590us/step

Testing for epoch 6 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 7 of 60

Testing for epoch 7 index 1:
32/32 [==============================] - 0s 773us/step

Testing for epoch 7 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 8 of 60

Testing for epoch 8 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 8 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 9 of 60

Testing for epoch 9 index 1:
32/32 [==============================] - 0s 599us/step

Testing for epoch 9 index 2:
32/32 [==============================] - 0s 571us/step
Epoch 10 of 60

Testing for epoch 10 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 10 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 11 of 60

Testing for epoch 11 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 11 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 12 of 60

Testing for epoch 12 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 12 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 13 of 60

Testing for epoch 13 index 1:
32/32 [==============================] - 0s 977us/step

Testing for epoch 13 index 2:
32/32 [==============================] - 0s 577us/step
Epoch 14 of 60

Testing for epoch 14 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 14 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 15 of 60

Testing for epoch 15 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 15 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 16 of 60

Testing for epoch 16 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 16 index 2:
32/32 [==============================] - 0s 577us/step
Epoch 17 of 60

Testing for epoch 17 index 1:
32/32 [==============================] - 0s 570us/step

Testing for epoch 17 index 2:
32/32 [==============================] - 0s 564us/step
Epoch 18 of 60

Testing for epoch 18 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 18 index 2:
32/32 [==============================] - 0s 737us/step
Epoch 19 of 60

Testing for epoch 19 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 19 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 20 of 60

Testing for epoch 20 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 20 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 21 of 60

Testing for epoch 21 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 21 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.4345
16/16 [==============================] - 0s 2ms/step - loss: 0.7133
16/16 [==============================] - 0s 2ms/step - loss: 0.9237
16/16 [==============================] - 0s 992us/step - loss: 1.0992
16/16 [==============================] - 0s 2ms/step - loss: 1.2285
16/16 [==============================] - 0s 2ms/step - loss: 1.2994
16/16 [==============================] - 0s 2ms/step - loss: 1.3332
16/16 [==============================] - 0s 2ms/step - loss: 1.3530
16/16 [==============================] - 0s 2ms/step - loss: 1.3618
16/16 [==============================] - 0s 844us/step - loss: 1.3663
Epoch 22 of 60

Testing for epoch 22 index 1:
32/32 [==============================] - 0s 566us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.4261
16/16 [==============================] - 0s 2ms/step - loss: 0.7204
16/16 [==============================] - 0s 2ms/step - loss: 0.9432
16/16 [==============================] - 0s 990us/step - loss: 1.1288
16/16 [==============================] - 0s 1ms/step - loss: 1.2626
16/16 [==============================] - 0s 2ms/step - loss: 1.3345
16/16 [==============================] - 0s 2ms/step - loss: 1.3678
16/16 [==============================] - 0s 1ms/step - loss: 1.3869
16/16 [==============================] - 0s 2ms/step - loss: 1.3953
16/16 [==============================] - 0s 2ms/step - loss: 1.3994

Testing for epoch 22 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 834us/step - loss: 0.4211
16/16 [==============================] - 0s 2ms/step - loss: 0.7245
16/16 [==============================] - 0s 1ms/step - loss: 0.9541
16/16 [==============================] - 0s 863us/step - loss: 1.1454
16/16 [==============================] - 0s 2ms/step - loss: 1.2804
16/16 [==============================] - 0s 868us/step - loss: 1.3515
16/16 [==============================] - 0s 2ms/step - loss: 1.3836
16/16 [==============================] - 0s 876us/step - loss: 1.4016
16/16 [==============================] - 0s 1ms/step - loss: 1.4093
16/16 [==============================] - 0s 901us/step - loss: 1.4131
Epoch 23 of 60

Testing for epoch 23 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.4209
16/16 [==============================] - 0s 2ms/step - loss: 0.7261
16/16 [==============================] - 0s 2ms/step - loss: 0.9649
16/16 [==============================] - 0s 2ms/step - loss: 1.1574
16/16 [==============================] - 0s 1ms/step - loss: 1.2891
16/16 [==============================] - 0s 2ms/step - loss: 1.3559
16/16 [==============================] - 0s 2ms/step - loss: 1.3848
16/16 [==============================] - 0s 868us/step - loss: 1.4003
16/16 [==============================] - 0s 826us/step - loss: 1.4068
16/16 [==============================] - 0s 850us/step - loss: 1.4100

Testing for epoch 23 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.4184
16/16 [==============================] - 0s 2ms/step - loss: 0.7292
16/16 [==============================] - 0s 2ms/step - loss: 0.9809
16/16 [==============================] - 0s 875us/step - loss: 1.1760
16/16 [==============================] - 0s 1ms/step - loss: 1.3069
16/16 [==============================] - 0s 2ms/step - loss: 1.3704
16/16 [==============================] - 0s 1ms/step - loss: 1.3970
16/16 [==============================] - 0s 947us/step - loss: 1.4107
16/16 [==============================] - 0s 2ms/step - loss: 1.4164
16/16 [==============================] - 0s 2ms/step - loss: 1.4191
Epoch 24 of 60

Testing for epoch 24 index 1:
32/32 [==============================] - 0s 575us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.4212
16/16 [==============================] - 0s 1ms/step - loss: 0.7335
16/16 [==============================] - 0s 2ms/step - loss: 0.9945
16/16 [==============================] - 0s 2ms/step - loss: 1.1886
16/16 [==============================] - 0s 988us/step - loss: 1.3164
16/16 [==============================] - 0s 2ms/step - loss: 1.3739
16/16 [==============================] - 0s 1ms/step - loss: 1.3967
16/16 [==============================] - 0s 2ms/step - loss: 1.4083
16/16 [==============================] - 0s 894us/step - loss: 1.4130
16/16 [==============================] - 0s 847us/step - loss: 1.4152

Testing for epoch 24 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.4112
16/16 [==============================] - 0s 826us/step - loss: 0.7397
16/16 [==============================] - 0s 2ms/step - loss: 1.0222
16/16 [==============================] - 0s 2ms/step - loss: 1.2268
16/16 [==============================] - 0s 2ms/step - loss: 1.3574
16/16 [==============================] - 0s 2ms/step - loss: 1.4137
16/16 [==============================] - 0s 914us/step - loss: 1.4346
16/16 [==============================] - 0s 1ms/step - loss: 1.4449
16/16 [==============================] - 0s 837us/step - loss: 1.4490
16/16 [==============================] - 0s 1ms/step - loss: 1.4509
Epoch 25 of 60

Testing for epoch 25 index 1:
32/32 [==============================] - 0s 690us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.4111
16/16 [==============================] - 0s 2ms/step - loss: 0.7434
16/16 [==============================] - 0s 2ms/step - loss: 1.0359
16/16 [==============================] - 0s 923us/step - loss: 1.2430
16/16 [==============================] - 0s 2ms/step - loss: 1.3672
16/16 [==============================] - 0s 2ms/step - loss: 1.4166
16/16 [==============================] - 0s 2ms/step - loss: 1.4345
16/16 [==============================] - 0s 836us/step - loss: 1.4430
16/16 [==============================] - 0s 2ms/step - loss: 1.4463
16/16 [==============================] - 0s 813us/step - loss: 1.4479

Testing for epoch 25 index 2:
32/32 [==============================] - 0s 578us/step
16/16 [==============================] - 0s 794us/step - loss: 0.4204
16/16 [==============================] - 0s 786us/step - loss: 0.7453
16/16 [==============================] - 0s 2ms/step - loss: 1.0374
16/16 [==============================] - 0s 2ms/step - loss: 1.2369
16/16 [==============================] - 0s 802us/step - loss: 1.3520
16/16 [==============================] - 0s 772us/step - loss: 1.3951
16/16 [==============================] - 0s 774us/step - loss: 1.4104
16/16 [==============================] - 0s 1ms/step - loss: 1.4173
16/16 [==============================] - 0s 805us/step - loss: 1.4201
16/16 [==============================] - 0s 868us/step - loss: 1.4214
Epoch 26 of 60

Testing for epoch 26 index 1:
32/32 [==============================] - 0s 566us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.4233
16/16 [==============================] - 0s 2ms/step - loss: 0.7493
16/16 [==============================] - 0s 2ms/step - loss: 1.0474
16/16 [==============================] - 0s 1ms/step - loss: 1.2468
16/16 [==============================] - 0s 1ms/step - loss: 1.3530
16/16 [==============================] - 0s 2ms/step - loss: 1.3921
16/16 [==============================] - 0s 839us/step - loss: 1.4053
16/16 [==============================] - 0s 2ms/step - loss: 1.4109
16/16 [==============================] - 0s 885us/step - loss: 1.4132
16/16 [==============================] - 0s 911us/step - loss: 1.4143

Testing for epoch 26 index 2:
32/32 [==============================] - 0s 569us/step
16/16 [==============================] - 0s 776us/step - loss: 0.4195
16/16 [==============================] - 0s 772us/step - loss: 0.7540
16/16 [==============================] - 0s 795us/step - loss: 1.0687
16/16 [==============================] - 0s 790us/step - loss: 1.2702
16/16 [==============================] - 0s 816us/step - loss: 1.3736
16/16 [==============================] - 0s 778us/step - loss: 1.4104
16/16 [==============================] - 0s 798us/step - loss: 1.4223
16/16 [==============================] - 0s 792us/step - loss: 1.4272
16/16 [==============================] - 0s 761us/step - loss: 1.4291
16/16 [==============================] - 0s 793us/step - loss: 1.4301
Epoch 27 of 60

Testing for epoch 27 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 781us/step - loss: 0.4196
16/16 [==============================] - 0s 807us/step - loss: 0.7599
16/16 [==============================] - 0s 781us/step - loss: 1.0875
16/16 [==============================] - 0s 775us/step - loss: 1.2839
16/16 [==============================] - 0s 785us/step - loss: 1.3836
16/16 [==============================] - 0s 2ms/step - loss: 1.4168
16/16 [==============================] - 0s 2ms/step - loss: 1.4273
16/16 [==============================] - 0s 877us/step - loss: 1.4314
16/16 [==============================] - 0s 824us/step - loss: 1.4330
16/16 [==============================] - 0s 785us/step - loss: 1.4339

Testing for epoch 27 index 2:
32/32 [==============================] - 0s 565us/step
16/16 [==============================] - 0s 779us/step - loss: 0.4253
16/16 [==============================] - 0s 781us/step - loss: 0.7594
16/16 [==============================] - 0s 776us/step - loss: 1.0874
16/16 [==============================] - 0s 2ms/step - loss: 1.2752
16/16 [==============================] - 0s 798us/step - loss: 1.3672
16/16 [==============================] - 0s 1ms/step - loss: 1.3966
16/16 [==============================] - 0s 2ms/step - loss: 1.4057
16/16 [==============================] - 0s 1ms/step - loss: 1.4092
16/16 [==============================] - 0s 2ms/step - loss: 1.4106
16/16 [==============================] - 0s 2ms/step - loss: 1.4114
Epoch 28 of 60

Testing for epoch 28 index 1:
32/32 [==============================] - 0s 556us/step
16/16 [==============================] - 0s 788us/step - loss: 0.4191
16/16 [==============================] - 0s 797us/step - loss: 0.7672
16/16 [==============================] - 0s 802us/step - loss: 1.1148
16/16 [==============================] - 0s 825us/step - loss: 1.3046
16/16 [==============================] - 0s 2ms/step - loss: 1.3941
16/16 [==============================] - 0s 834us/step - loss: 1.4213
16/16 [==============================] - 0s 802us/step - loss: 1.4294
16/16 [==============================] - 0s 2ms/step - loss: 1.4325
16/16 [==============================] - 0s 2ms/step - loss: 1.4337
16/16 [==============================] - 0s 806us/step - loss: 1.4345

Testing for epoch 28 index 2:
32/32 [==============================] - 0s 577us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.4230
16/16 [==============================] - 0s 1ms/step - loss: 0.7713
16/16 [==============================] - 0s 799us/step - loss: 1.1226
16/16 [==============================] - 0s 787us/step - loss: 1.3086
16/16 [==============================] - 0s 778us/step - loss: 1.3931
16/16 [==============================] - 0s 802us/step - loss: 1.4179
16/16 [==============================] - 0s 788us/step - loss: 1.4251
16/16 [==============================] - 0s 2ms/step - loss: 1.4278
16/16 [==============================] - 0s 834us/step - loss: 1.4288
16/16 [==============================] - 0s 788us/step - loss: 1.4296
Epoch 29 of 60

Testing for epoch 29 index 1:
32/32 [==============================] - 0s 661us/step
16/16 [==============================] - 0s 822us/step - loss: 0.4148
16/16 [==============================] - 0s 789us/step - loss: 0.7758
16/16 [==============================] - 0s 1ms/step - loss: 1.1424
16/16 [==============================] - 0s 793us/step - loss: 1.3333
16/16 [==============================] - 0s 774us/step - loss: 1.4154
16/16 [==============================] - 0s 842us/step - loss: 1.4387
16/16 [==============================] - 0s 2ms/step - loss: 1.4453
16/16 [==============================] - 0s 1ms/step - loss: 1.4477
16/16 [==============================] - 0s 2ms/step - loss: 1.4486
16/16 [==============================] - 0s 2ms/step - loss: 1.4493

Testing for epoch 29 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.4225
16/16 [==============================] - 0s 2ms/step - loss: 0.7755
16/16 [==============================] - 0s 1ms/step - loss: 1.1337
16/16 [==============================] - 0s 871us/step - loss: 1.3172
16/16 [==============================] - 0s 793us/step - loss: 1.3930
16/16 [==============================] - 0s 779us/step - loss: 1.4138
16/16 [==============================] - 0s 777us/step - loss: 1.4195
16/16 [==============================] - 0s 2ms/step - loss: 1.4217
16/16 [==============================] - 0s 837us/step - loss: 1.4225
16/16 [==============================] - 0s 2ms/step - loss: 1.4232
Epoch 30 of 60

Testing for epoch 30 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.4221
16/16 [==============================] - 0s 2ms/step - loss: 0.7804
16/16 [==============================] - 0s 2ms/step - loss: 1.1445
16/16 [==============================] - 0s 1ms/step - loss: 1.3280
16/16 [==============================] - 0s 2ms/step - loss: 1.4003
16/16 [==============================] - 0s 2ms/step - loss: 1.4195
16/16 [==============================] - 0s 2ms/step - loss: 1.4246
16/16 [==============================] - 0s 2ms/step - loss: 1.4266
16/16 [==============================] - 0s 2ms/step - loss: 1.4274
16/16 [==============================] - 0s 837us/step - loss: 1.4281

Testing for epoch 30 index 2:
32/32 [==============================] - 0s 555us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.4129
16/16 [==============================] - 0s 1ms/step - loss: 0.7881
16/16 [==============================] - 0s 2ms/step - loss: 1.1734
16/16 [==============================] - 0s 2ms/step - loss: 1.3635
16/16 [==============================] - 0s 1ms/step - loss: 1.4366
16/16 [==============================] - 0s 2ms/step - loss: 1.4555
16/16 [==============================] - 0s 2ms/step - loss: 1.4605
16/16 [==============================] - 0s 2ms/step - loss: 1.4623
16/16 [==============================] - 0s 1ms/step - loss: 1.4630
16/16 [==============================] - 0s 1ms/step - loss: 1.4637
Epoch 31 of 60

Testing for epoch 31 index 1:
32/32 [==============================] - 0s 604us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.4140
16/16 [==============================] - 0s 998us/step - loss: 0.7917
16/16 [==============================] - 0s 2ms/step - loss: 1.1828
16/16 [==============================] - 0s 1ms/step - loss: 1.3694
16/16 [==============================] - 0s 1ms/step - loss: 1.4394
16/16 [==============================] - 0s 823us/step - loss: 1.4570
16/16 [==============================] - 0s 2ms/step - loss: 1.4615
16/16 [==============================] - 0s 2ms/step - loss: 1.4631
16/16 [==============================] - 0s 825us/step - loss: 1.4638
16/16 [==============================] - 0s 2ms/step - loss: 1.4645

Testing for epoch 31 index 2:
32/32 [==============================] - 0s 568us/step
16/16 [==============================] - 0s 797us/step - loss: 0.4160
16/16 [==============================] - 0s 2ms/step - loss: 0.7876
16/16 [==============================] - 0s 2ms/step - loss: 1.1724
16/16 [==============================] - 0s 2ms/step - loss: 1.3533
16/16 [==============================] - 0s 864us/step - loss: 1.4196
16/16 [==============================] - 0s 968us/step - loss: 1.4358
16/16 [==============================] - 0s 2ms/step - loss: 1.4400
16/16 [==============================] - 0s 873us/step - loss: 1.4415
16/16 [==============================] - 0s 791us/step - loss: 1.4422
16/16 [==============================] - 0s 794us/step - loss: 1.4428
Epoch 32 of 60

Testing for epoch 32 index 1:
32/32 [==============================] - 0s 583us/step
16/16 [==============================] - 0s 788us/step - loss: 0.4174
16/16 [==============================] - 0s 2ms/step - loss: 0.7933
16/16 [==============================] - 0s 2ms/step - loss: 1.1831
16/16 [==============================] - 0s 2ms/step - loss: 1.3641
16/16 [==============================] - 0s 862us/step - loss: 1.4285
16/16 [==============================] - 0s 770us/step - loss: 1.4438
16/16 [==============================] - 0s 794us/step - loss: 1.4477
16/16 [==============================] - 0s 754us/step - loss: 1.4491
16/16 [==============================] - 0s 2ms/step - loss: 1.4497
16/16 [==============================] - 0s 785us/step - loss: 1.4504

Testing for epoch 32 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 782us/step - loss: 0.4087
16/16 [==============================] - 0s 2ms/step - loss: 0.7935
16/16 [==============================] - 0s 964us/step - loss: 1.1935
16/16 [==============================] - 0s 2ms/step - loss: 1.3777
16/16 [==============================] - 0s 831us/step - loss: 1.4420
16/16 [==============================] - 0s 2ms/step - loss: 1.4570
16/16 [==============================] - 0s 1ms/step - loss: 1.4608
16/16 [==============================] - 0s 2ms/step - loss: 1.4621
16/16 [==============================] - 0s 930us/step - loss: 1.4628
16/16 [==============================] - 0s 1ms/step - loss: 1.4635
Epoch 33 of 60

Testing for epoch 33 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 841us/step - loss: 0.4160
16/16 [==============================] - 0s 2ms/step - loss: 0.7925
16/16 [==============================] - 0s 2ms/step - loss: 1.1814
16/16 [==============================] - 0s 2ms/step - loss: 1.3582
16/16 [==============================] - 0s 1ms/step - loss: 1.4186
16/16 [==============================] - 0s 2ms/step - loss: 1.4324
16/16 [==============================] - 0s 853us/step - loss: 1.4358
16/16 [==============================] - 0s 786us/step - loss: 1.4370
16/16 [==============================] - 0s 788us/step - loss: 1.4377
16/16 [==============================] - 0s 798us/step - loss: 1.4384

Testing for epoch 33 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 872us/step - loss: 0.4056
16/16 [==============================] - 0s 2ms/step - loss: 0.8016
16/16 [==============================] - 0s 800us/step - loss: 1.2083
16/16 [==============================] - 0s 2ms/step - loss: 1.3936
16/16 [==============================] - 0s 1ms/step - loss: 1.4561
16/16 [==============================] - 0s 2ms/step - loss: 1.4702
16/16 [==============================] - 0s 801us/step - loss: 1.4736
16/16 [==============================] - 0s 2ms/step - loss: 1.4749
16/16 [==============================] - 0s 817us/step - loss: 1.4755
16/16 [==============================] - 0s 2ms/step - loss: 1.4762
Epoch 34 of 60

Testing for epoch 34 index 1:
32/32 [==============================] - 0s 563us/step
16/16 [==============================] - 0s 802us/step - loss: 0.3972
16/16 [==============================] - 0s 813us/step - loss: 0.8100
16/16 [==============================] - 0s 793us/step - loss: 1.2301
16/16 [==============================] - 0s 2ms/step - loss: 1.4208
16/16 [==============================] - 0s 806us/step - loss: 1.4840
16/16 [==============================] - 0s 1ms/step - loss: 1.4980
16/16 [==============================] - 0s 765us/step - loss: 1.5014
16/16 [==============================] - 0s 786us/step - loss: 1.5026
16/16 [==============================] - 0s 775us/step - loss: 1.5032
16/16 [==============================] - 0s 1ms/step - loss: 1.5039

Testing for epoch 34 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.3993
16/16 [==============================] - 0s 2ms/step - loss: 0.8090
16/16 [==============================] - 0s 865us/step - loss: 1.2207
16/16 [==============================] - 0s 846us/step - loss: 1.4083
16/16 [==============================] - 0s 837us/step - loss: 1.4698
16/16 [==============================] - 0s 2ms/step - loss: 1.4832
16/16 [==============================] - 0s 816us/step - loss: 1.4864
16/16 [==============================] - 0s 2ms/step - loss: 1.4876
16/16 [==============================] - 0s 915us/step - loss: 1.4882
16/16 [==============================] - 0s 844us/step - loss: 1.4889
Epoch 35 of 60

Testing for epoch 35 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 858us/step - loss: 0.3885
16/16 [==============================] - 0s 2ms/step - loss: 0.8232
16/16 [==============================] - 0s 2ms/step - loss: 1.2556
16/16 [==============================] - 0s 801us/step - loss: 1.4530
16/16 [==============================] - 0s 2ms/step - loss: 1.5168
16/16 [==============================] - 0s 2ms/step - loss: 1.5306
16/16 [==============================] - 0s 2ms/step - loss: 1.5339
16/16 [==============================] - 0s 806us/step - loss: 1.5350
16/16 [==============================] - 0s 905us/step - loss: 1.5356
16/16 [==============================] - 0s 2ms/step - loss: 1.5363

Testing for epoch 35 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 885us/step - loss: 0.3917
16/16 [==============================] - 0s 815us/step - loss: 0.8212
16/16 [==============================] - 0s 822us/step - loss: 1.2414
16/16 [==============================] - 0s 921us/step - loss: 1.4339
16/16 [==============================] - 0s 933us/step - loss: 1.4957
16/16 [==============================] - 0s 841us/step - loss: 1.5090
16/16 [==============================] - 0s 799us/step - loss: 1.5121
16/16 [==============================] - 0s 786us/step - loss: 1.5132
16/16 [==============================] - 0s 1ms/step - loss: 1.5138
16/16 [==============================] - 0s 801us/step - loss: 1.5145
Epoch 36 of 60

Testing for epoch 36 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.3811
16/16 [==============================] - 0s 1ms/step - loss: 0.8277
16/16 [==============================] - 0s 2ms/step - loss: 1.2615
16/16 [==============================] - 0s 814us/step - loss: 1.4581
16/16 [==============================] - 0s 886us/step - loss: 1.5210
16/16 [==============================] - 0s 777us/step - loss: 1.5343
16/16 [==============================] - 0s 792us/step - loss: 1.5374
16/16 [==============================] - 0s 760us/step - loss: 1.5385
16/16 [==============================] - 0s 785us/step - loss: 1.5391
16/16 [==============================] - 0s 775us/step - loss: 1.5398

Testing for epoch 36 index 2:
32/32 [==============================] - 0s 574us/step
16/16 [==============================] - 0s 766us/step - loss: 0.3871
16/16 [==============================] - 0s 794us/step - loss: 0.8338
16/16 [==============================] - 0s 928us/step - loss: 1.2599
16/16 [==============================] - 0s 1ms/step - loss: 1.4551
16/16 [==============================] - 0s 1ms/step - loss: 1.5166
16/16 [==============================] - 0s 2ms/step - loss: 1.5297
16/16 [==============================] - 0s 2ms/step - loss: 1.5327
16/16 [==============================] - 0s 794us/step - loss: 1.5338
16/16 [==============================] - 0s 838us/step - loss: 1.5344
16/16 [==============================] - 0s 2ms/step - loss: 1.5351
Epoch 37 of 60

Testing for epoch 37 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 807us/step - loss: 0.3797
16/16 [==============================] - 0s 812us/step - loss: 0.8402
16/16 [==============================] - 0s 778us/step - loss: 1.2755
16/16 [==============================] - 0s 789us/step - loss: 1.4745
16/16 [==============================] - 0s 800us/step - loss: 1.5370
16/16 [==============================] - 0s 798us/step - loss: 1.5501
16/16 [==============================] - 0s 795us/step - loss: 1.5531
16/16 [==============================] - 0s 785us/step - loss: 1.5542
16/16 [==============================] - 0s 767us/step - loss: 1.5548
16/16 [==============================] - 0s 864us/step - loss: 1.5555

Testing for epoch 37 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 839us/step - loss: 0.3801
16/16 [==============================] - 0s 786us/step - loss: 0.8431
16/16 [==============================] - 0s 774us/step - loss: 1.2754
16/16 [==============================] - 0s 793us/step - loss: 1.4705
16/16 [==============================] - 0s 790us/step - loss: 1.5327
16/16 [==============================] - 0s 2ms/step - loss: 1.5457
16/16 [==============================] - 0s 2ms/step - loss: 1.5487
16/16 [==============================] - 0s 2ms/step - loss: 1.5497
16/16 [==============================] - 0s 979us/step - loss: 1.5504
16/16 [==============================] - 0s 906us/step - loss: 1.5511
Epoch 38 of 60

Testing for epoch 38 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 799us/step - loss: 0.3624
16/16 [==============================] - 0s 776us/step - loss: 0.8601
16/16 [==============================] - 0s 2ms/step - loss: 1.3194
16/16 [==============================] - 0s 2ms/step - loss: 1.5287
16/16 [==============================] - 0s 1ms/step - loss: 1.5949
16/16 [==============================] - 0s 2ms/step - loss: 1.6086
16/16 [==============================] - 0s 1ms/step - loss: 1.6117
16/16 [==============================] - 0s 2ms/step - loss: 1.6128
16/16 [==============================] - 0s 877us/step - loss: 1.6134
16/16 [==============================] - 0s 847us/step - loss: 1.6141

Testing for epoch 38 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 847us/step - loss: 0.3678
16/16 [==============================] - 0s 793us/step - loss: 0.8486
16/16 [==============================] - 0s 839us/step - loss: 1.2889
16/16 [==============================] - 0s 898us/step - loss: 1.4911
16/16 [==============================] - 0s 2ms/step - loss: 1.5547
16/16 [==============================] - 0s 1ms/step - loss: 1.5678
16/16 [==============================] - 0s 2ms/step - loss: 1.5708
16/16 [==============================] - 0s 2ms/step - loss: 1.5719
16/16 [==============================] - 0s 2ms/step - loss: 1.5725
16/16 [==============================] - 0s 944us/step - loss: 1.5733
Epoch 39 of 60

Testing for epoch 39 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 829us/step - loss: 0.3676
16/16 [==============================] - 0s 2ms/step - loss: 0.8484
16/16 [==============================] - 0s 2ms/step - loss: 1.2921
16/16 [==============================] - 0s 2ms/step - loss: 1.4923
16/16 [==============================] - 0s 908us/step - loss: 1.5556
16/16 [==============================] - 0s 952us/step - loss: 1.5685
16/16 [==============================] - 0s 1ms/step - loss: 1.5714
16/16 [==============================] - 0s 862us/step - loss: 1.5725
16/16 [==============================] - 0s 811us/step - loss: 1.5731
16/16 [==============================] - 0s 1ms/step - loss: 1.5739

Testing for epoch 39 index 2:
32/32 [==============================] - 0s 568us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.3635
16/16 [==============================] - 0s 2ms/step - loss: 0.8481
16/16 [==============================] - 0s 2ms/step - loss: 1.2943
16/16 [==============================] - 0s 2ms/step - loss: 1.4972
16/16 [==============================] - 0s 921us/step - loss: 1.5612
16/16 [==============================] - 0s 1ms/step - loss: 1.5742
16/16 [==============================] - 0s 2ms/step - loss: 1.5772
16/16 [==============================] - 0s 873us/step - loss: 1.5782
16/16 [==============================] - 0s 2ms/step - loss: 1.5789
16/16 [==============================] - 0s 815us/step - loss: 1.5796
Epoch 40 of 60

Testing for epoch 40 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 799us/step - loss: 0.3622
16/16 [==============================] - 0s 2ms/step - loss: 0.8582
16/16 [==============================] - 0s 2ms/step - loss: 1.3106
16/16 [==============================] - 0s 2ms/step - loss: 1.5171
16/16 [==============================] - 0s 830us/step - loss: 1.5820
16/16 [==============================] - 0s 2ms/step - loss: 1.5952
16/16 [==============================] - 0s 1ms/step - loss: 1.5981
16/16 [==============================] - 0s 2ms/step - loss: 1.5991
16/16 [==============================] - 0s 817us/step - loss: 1.5997
16/16 [==============================] - 0s 769us/step - loss: 1.6004

Testing for epoch 40 index 2:
32/32 [==============================] - 0s 785us/step
16/16 [==============================] - 0s 804us/step - loss: 0.3578
16/16 [==============================] - 0s 2ms/step - loss: 0.8585
16/16 [==============================] - 0s 2ms/step - loss: 1.3090
16/16 [==============================] - 0s 869us/step - loss: 1.5148
16/16 [==============================] - 0s 826us/step - loss: 1.5799
16/16 [==============================] - 0s 2ms/step - loss: 1.5931
16/16 [==============================] - 0s 2ms/step - loss: 1.5960
16/16 [==============================] - 0s 887us/step - loss: 1.5971
16/16 [==============================] - 0s 2ms/step - loss: 1.5977
16/16 [==============================] - 0s 929us/step - loss: 1.5985
Epoch 41 of 60

Testing for epoch 41 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.3466
16/16 [==============================] - 0s 2ms/step - loss: 0.8702
16/16 [==============================] - 0s 868us/step - loss: 1.3373
16/16 [==============================] - 0s 2ms/step - loss: 1.5541
16/16 [==============================] - 0s 2ms/step - loss: 1.6203
16/16 [==============================] - 0s 874us/step - loss: 1.6339
16/16 [==============================] - 0s 2ms/step - loss: 1.6369
16/16 [==============================] - 0s 1ms/step - loss: 1.6380
16/16 [==============================] - 0s 2ms/step - loss: 1.6386
16/16 [==============================] - 0s 891us/step - loss: 1.6393

Testing for epoch 41 index 2:
32/32 [==============================] - 0s 571us/step
16/16 [==============================] - 0s 1ms/step - loss: 0.3596
16/16 [==============================] - 0s 2ms/step - loss: 0.8626
16/16 [==============================] - 0s 2ms/step - loss: 1.3030
16/16 [==============================] - 0s 839us/step - loss: 1.5093
16/16 [==============================] - 0s 834us/step - loss: 1.5721
16/16 [==============================] - 0s 773us/step - loss: 1.5851
16/16 [==============================] - 0s 805us/step - loss: 1.5879
16/16 [==============================] - 0s 2ms/step - loss: 1.5889
16/16 [==============================] - 0s 1ms/step - loss: 1.5896
16/16 [==============================] - 0s 2ms/step - loss: 1.5904
Epoch 42 of 60

Testing for epoch 42 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.3466
16/16 [==============================] - 0s 1ms/step - loss: 0.8730
16/16 [==============================] - 0s 2ms/step - loss: 1.3386
16/16 [==============================] - 0s 904us/step - loss: 1.5564
16/16 [==============================] - 0s 842us/step - loss: 1.6225
16/16 [==============================] - 0s 797us/step - loss: 1.6360
16/16 [==============================] - 0s 790us/step - loss: 1.6390
16/16 [==============================] - 0s 776us/step - loss: 1.6400
16/16 [==============================] - 0s 802us/step - loss: 1.6406
16/16 [==============================] - 0s 797us/step - loss: 1.6414

Testing for epoch 42 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 812us/step - loss: 0.3347
16/16 [==============================] - 0s 793us/step - loss: 0.8792
16/16 [==============================] - 0s 811us/step - loss: 1.3648
16/16 [==============================] - 0s 820us/step - loss: 1.5920
16/16 [==============================] - 0s 837us/step - loss: 1.6609
16/16 [==============================] - 0s 2ms/step - loss: 1.6751
16/16 [==============================] - 0s 2ms/step - loss: 1.6781
16/16 [==============================] - 0s 823us/step - loss: 1.6792
16/16 [==============================] - 0s 759us/step - loss: 1.6798
16/16 [==============================] - 0s 807us/step - loss: 1.6806
Epoch 43 of 60

Testing for epoch 43 index 1:
32/32 [==============================] - 0s 680us/step
16/16 [==============================] - 0s 784us/step - loss: 0.3423
16/16 [==============================] - 0s 783us/step - loss: 0.8698
16/16 [==============================] - 0s 917us/step - loss: 1.3426
16/16 [==============================] - 0s 942us/step - loss: 1.5630
16/16 [==============================] - 0s 948us/step - loss: 1.6294
16/16 [==============================] - 0s 2ms/step - loss: 1.6429
16/16 [==============================] - 0s 2ms/step - loss: 1.6458
16/16 [==============================] - 0s 1ms/step - loss: 1.6469
16/16 [==============================] - 0s 938us/step - loss: 1.6475
16/16 [==============================] - 0s 1ms/step - loss: 1.6483

Testing for epoch 43 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.3489
16/16 [==============================] - 0s 1ms/step - loss: 0.8556
16/16 [==============================] - 0s 780us/step - loss: 1.3105
16/16 [==============================] - 0s 828us/step - loss: 1.5222
16/16 [==============================] - 0s 931us/step - loss: 1.5859
16/16 [==============================] - 0s 2ms/step - loss: 1.5988
16/16 [==============================] - 0s 1ms/step - loss: 1.6016
16/16 [==============================] - 0s 2ms/step - loss: 1.6026
16/16 [==============================] - 0s 850us/step - loss: 1.6033
16/16 [==============================] - 0s 2ms/step - loss: 1.6041
Epoch 44 of 60

Testing for epoch 44 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.3334
16/16 [==============================] - 0s 1ms/step - loss: 0.8738
16/16 [==============================] - 0s 1ms/step - loss: 1.3626
16/16 [==============================] - 0s 882us/step - loss: 1.5900
16/16 [==============================] - 0s 859us/step - loss: 1.6580
16/16 [==============================] - 0s 2ms/step - loss: 1.6717
16/16 [==============================] - 0s 2ms/step - loss: 1.6746
16/16 [==============================] - 0s 2ms/step - loss: 1.6757
16/16 [==============================] - 0s 2ms/step - loss: 1.6763
16/16 [==============================] - 0s 1ms/step - loss: 1.6771

Testing for epoch 44 index 2:
32/32 [==============================] - 0s 553us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.3280
16/16 [==============================] - 0s 915us/step - loss: 0.8719
16/16 [==============================] - 0s 2ms/step - loss: 1.3668
16/16 [==============================] - 0s 756us/step - loss: 1.5971
16/16 [==============================] - 0s 791us/step - loss: 1.6658
16/16 [==============================] - 0s 781us/step - loss: 1.6796
16/16 [==============================] - 0s 809us/step - loss: 1.6825
16/16 [==============================] - 0s 804us/step - loss: 1.6835
16/16 [==============================] - 0s 1ms/step - loss: 1.6842
16/16 [==============================] - 0s 1ms/step - loss: 1.6850
Epoch 45 of 60

Testing for epoch 45 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 805us/step - loss: 0.3312
16/16 [==============================] - 0s 2ms/step - loss: 0.8689
16/16 [==============================] - 0s 778us/step - loss: 1.3585
16/16 [==============================] - 0s 781us/step - loss: 1.5852
16/16 [==============================] - 0s 793us/step - loss: 1.6522
16/16 [==============================] - 0s 777us/step - loss: 1.6655
16/16 [==============================] - 0s 1ms/step - loss: 1.6683
16/16 [==============================] - 0s 792us/step - loss: 1.6693
16/16 [==============================] - 0s 778us/step - loss: 1.6700
16/16 [==============================] - 0s 1ms/step - loss: 1.6708

Testing for epoch 45 index 2:
32/32 [==============================] - 0s 617us/step
16/16 [==============================] - 0s 803us/step - loss: 0.3327
16/16 [==============================] - 0s 2ms/step - loss: 0.8664
16/16 [==============================] - 0s 2ms/step - loss: 1.3547
16/16 [==============================] - 0s 845us/step - loss: 1.5807
16/16 [==============================] - 0s 766us/step - loss: 1.6474
16/16 [==============================] - 0s 2ms/step - loss: 1.6606
16/16 [==============================] - 0s 2ms/step - loss: 1.6634
16/16 [==============================] - 0s 2ms/step - loss: 1.6644
16/16 [==============================] - 0s 903us/step - loss: 1.6650
16/16 [==============================] - 0s 2ms/step - loss: 1.6659
Epoch 46 of 60

Testing for epoch 46 index 1:
32/32 [==============================] - 0s 780us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.3185
16/16 [==============================] - 0s 2ms/step - loss: 0.8798
16/16 [==============================] - 0s 1ms/step - loss: 1.3925
16/16 [==============================] - 0s 2ms/step - loss: 1.6304
16/16 [==============================] - 0s 760us/step - loss: 1.7001
16/16 [==============================] - 0s 2ms/step - loss: 1.7138
16/16 [==============================] - 0s 2ms/step - loss: 1.7166
16/16 [==============================] - 0s 810us/step - loss: 1.7176
16/16 [==============================] - 0s 788us/step - loss: 1.7183
16/16 [==============================] - 0s 1ms/step - loss: 1.7191

Testing for epoch 46 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 843us/step - loss: 0.3095
16/16 [==============================] - 0s 2ms/step - loss: 0.8859
16/16 [==============================] - 0s 835us/step - loss: 1.4168
16/16 [==============================] - 0s 802us/step - loss: 1.6631
16/16 [==============================] - 0s 1ms/step - loss: 1.7351
16/16 [==============================] - 0s 882us/step - loss: 1.7492
16/16 [==============================] - 0s 912us/step - loss: 1.7521
16/16 [==============================] - 0s 862us/step - loss: 1.7531
16/16 [==============================] - 0s 803us/step - loss: 1.7538
16/16 [==============================] - 0s 2ms/step - loss: 1.7545
Epoch 47 of 60

Testing for epoch 47 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.3166
16/16 [==============================] - 0s 2ms/step - loss: 0.8807
16/16 [==============================] - 0s 2ms/step - loss: 1.4012
16/16 [==============================] - 0s 2ms/step - loss: 1.6415
16/16 [==============================] - 0s 2ms/step - loss: 1.7109
16/16 [==============================] - 0s 2ms/step - loss: 1.7244
16/16 [==============================] - 0s 902us/step - loss: 1.7272
16/16 [==============================] - 0s 1ms/step - loss: 1.7282
16/16 [==============================] - 0s 1ms/step - loss: 1.7288
16/16 [==============================] - 0s 2ms/step - loss: 1.7296

Testing for epoch 47 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.3071
16/16 [==============================] - 0s 2ms/step - loss: 0.8856
16/16 [==============================] - 0s 1ms/step - loss: 1.4225
16/16 [==============================] - 0s 941us/step - loss: 1.6700
16/16 [==============================] - 0s 2ms/step - loss: 1.7412
16/16 [==============================] - 0s 815us/step - loss: 1.7550
16/16 [==============================] - 0s 769us/step - loss: 1.7578
16/16 [==============================] - 0s 809us/step - loss: 1.7588
16/16 [==============================] - 0s 932us/step - loss: 1.7595
16/16 [==============================] - 0s 2ms/step - loss: 1.7603
Epoch 48 of 60

Testing for epoch 48 index 1:
32/32 [==============================] - 0s 573us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.3016
16/16 [==============================] - 0s 2ms/step - loss: 0.8889
16/16 [==============================] - 0s 1ms/step - loss: 1.4374
16/16 [==============================] - 0s 1ms/step - loss: 1.6892
16/16 [==============================] - 0s 786us/step - loss: 1.7611
16/16 [==============================] - 0s 787us/step - loss: 1.7748
16/16 [==============================] - 0s 901us/step - loss: 1.7776
16/16 [==============================] - 0s 978us/step - loss: 1.7786
16/16 [==============================] - 0s 2ms/step - loss: 1.7792
16/16 [==============================] - 0s 817us/step - loss: 1.7800

Testing for epoch 48 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.3078
16/16 [==============================] - 0s 2ms/step - loss: 0.8793
16/16 [==============================] - 0s 2ms/step - loss: 1.4145
16/16 [==============================] - 0s 2ms/step - loss: 1.6595
16/16 [==============================] - 0s 837us/step - loss: 1.7291
16/16 [==============================] - 0s 1ms/step - loss: 1.7423
16/16 [==============================] - 0s 889us/step - loss: 1.7449
16/16 [==============================] - 0s 833us/step - loss: 1.7459
16/16 [==============================] - 0s 777us/step - loss: 1.7466
16/16 [==============================] - 0s 2ms/step - loss: 1.7474
Epoch 49 of 60

Testing for epoch 49 index 1:
32/32 [==============================] - 0s 578us/step
16/16 [==============================] - 0s 902us/step - loss: 0.3141
16/16 [==============================] - 0s 2ms/step - loss: 0.8736
16/16 [==============================] - 0s 2ms/step - loss: 1.3976
16/16 [==============================] - 0s 2ms/step - loss: 1.6358
16/16 [==============================] - 0s 880us/step - loss: 1.7026
16/16 [==============================] - 0s 955us/step - loss: 1.7151
16/16 [==============================] - 0s 1ms/step - loss: 1.7176
16/16 [==============================] - 0s 1ms/step - loss: 1.7186
16/16 [==============================] - 0s 802us/step - loss: 1.7192
16/16 [==============================] - 0s 2ms/step - loss: 1.7201

Testing for epoch 49 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.2951
16/16 [==============================] - 0s 833us/step - loss: 0.8939
16/16 [==============================] - 0s 823us/step - loss: 1.4599
16/16 [==============================] - 0s 2ms/step - loss: 1.7171
16/16 [==============================] - 0s 2ms/step - loss: 1.7890
16/16 [==============================] - 0s 2ms/step - loss: 1.8024
16/16 [==============================] - 0s 2ms/step - loss: 1.8051
16/16 [==============================] - 0s 921us/step - loss: 1.8060
16/16 [==============================] - 0s 863us/step - loss: 1.8067
16/16 [==============================] - 0s 787us/step - loss: 1.8075
Epoch 50 of 60

Testing for epoch 50 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 793us/step - loss: 0.3033
16/16 [==============================] - 0s 1ms/step - loss: 0.8913
16/16 [==============================] - 0s 815us/step - loss: 1.4487
16/16 [==============================] - 0s 786us/step - loss: 1.7005
16/16 [==============================] - 0s 770us/step - loss: 1.7700
16/16 [==============================] - 0s 807us/step - loss: 1.7828
16/16 [==============================] - 0s 954us/step - loss: 1.7853
16/16 [==============================] - 0s 1ms/step - loss: 1.7862
16/16 [==============================] - 0s 790us/step - loss: 1.7868
16/16 [==============================] - 0s 769us/step - loss: 1.7876

Testing for epoch 50 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 862us/step - loss: 0.2983
16/16 [==============================] - 0s 2ms/step - loss: 0.8872
16/16 [==============================] - 0s 794us/step - loss: 1.4445
16/16 [==============================] - 0s 1ms/step - loss: 1.6957
16/16 [==============================] - 0s 981us/step - loss: 1.7645
16/16 [==============================] - 0s 792us/step - loss: 1.7771
16/16 [==============================] - 0s 872us/step - loss: 1.7796
16/16 [==============================] - 0s 774us/step - loss: 1.7805
16/16 [==============================] - 0s 2ms/step - loss: 1.7811
16/16 [==============================] - 0s 1ms/step - loss: 1.7820
Epoch 51 of 60

Testing for epoch 51 index 1:
32/32 [==============================] - 0s 580us/step
16/16 [==============================] - 0s 809us/step - loss: 0.2901
16/16 [==============================] - 0s 2ms/step - loss: 0.8941
16/16 [==============================] - 0s 821us/step - loss: 1.4669
16/16 [==============================] - 0s 810us/step - loss: 1.7233
16/16 [==============================] - 0s 2ms/step - loss: 1.7926
16/16 [==============================] - 0s 2ms/step - loss: 1.8051
16/16 [==============================] - 0s 1ms/step - loss: 1.8075
16/16 [==============================] - 0s 793us/step - loss: 1.8084
16/16 [==============================] - 0s 818us/step - loss: 1.8091
16/16 [==============================] - 0s 788us/step - loss: 1.8099

Testing for epoch 51 index 2:
32/32 [==============================] - 0s 2ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.2951
16/16 [==============================] - 0s 1ms/step - loss: 0.8865
16/16 [==============================] - 0s 1ms/step - loss: 1.4489
16/16 [==============================] - 0s 2ms/step - loss: 1.6998
16/16 [==============================] - 0s 947us/step - loss: 1.7671
16/16 [==============================] - 0s 1ms/step - loss: 1.7791
16/16 [==============================] - 0s 797us/step - loss: 1.7814
16/16 [==============================] - 0s 2ms/step - loss: 1.7823
16/16 [==============================] - 0s 2ms/step - loss: 1.7830
16/16 [==============================] - 0s 997us/step - loss: 1.7839
Epoch 52 of 60

Testing for epoch 52 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 793us/step - loss: 0.2874
16/16 [==============================] - 0s 920us/step - loss: 0.8992
16/16 [==============================] - 0s 2ms/step - loss: 1.4836
16/16 [==============================] - 0s 1ms/step - loss: 1.7430
16/16 [==============================] - 0s 2ms/step - loss: 1.8119
16/16 [==============================] - 0s 842us/step - loss: 1.8240
16/16 [==============================] - 0s 2ms/step - loss: 1.8263
16/16 [==============================] - 0s 2ms/step - loss: 1.8272
16/16 [==============================] - 0s 849us/step - loss: 1.8278
16/16 [==============================] - 0s 810us/step - loss: 1.8287

Testing for epoch 52 index 2:
32/32 [==============================] - 0s 663us/step
16/16 [==============================] - 0s 798us/step - loss: 0.2935
16/16 [==============================] - 0s 2ms/step - loss: 0.8916
16/16 [==============================] - 0s 2ms/step - loss: 1.4652
16/16 [==============================] - 0s 810us/step - loss: 1.7180
16/16 [==============================] - 0s 2ms/step - loss: 1.7850
16/16 [==============================] - 0s 1ms/step - loss: 1.7965
16/16 [==============================] - 0s 2ms/step - loss: 1.7987
16/16 [==============================] - 0s 2ms/step - loss: 1.7996
16/16 [==============================] - 0s 2ms/step - loss: 1.8002
16/16 [==============================] - 0s 916us/step - loss: 1.8011
Epoch 53 of 60

Testing for epoch 53 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 814us/step - loss: 0.2903
16/16 [==============================] - 0s 2ms/step - loss: 0.8924
16/16 [==============================] - 0s 1ms/step - loss: 1.4727
16/16 [==============================] - 0s 2ms/step - loss: 1.7236
16/16 [==============================] - 0s 2ms/step - loss: 1.7899
16/16 [==============================] - 0s 2ms/step - loss: 1.8009
16/16 [==============================] - 0s 1ms/step - loss: 1.8031
16/16 [==============================] - 0s 840us/step - loss: 1.8039
16/16 [==============================] - 0s 2ms/step - loss: 1.8046
16/16 [==============================] - 0s 2ms/step - loss: 1.8054

Testing for epoch 53 index 2:
32/32 [==============================] - 0s 626us/step
16/16 [==============================] - 0s 774us/step - loss: 0.2955
16/16 [==============================] - 0s 2ms/step - loss: 0.8898
16/16 [==============================] - 0s 1ms/step - loss: 1.4673
16/16 [==============================] - 0s 2ms/step - loss: 1.7113
16/16 [==============================] - 0s 2ms/step - loss: 1.7764
16/16 [==============================] - 0s 2ms/step - loss: 1.7869
16/16 [==============================] - 0s 944us/step - loss: 1.7890
16/16 [==============================] - 0s 2ms/step - loss: 1.7898
16/16 [==============================] - 0s 892us/step - loss: 1.7905
16/16 [==============================] - 0s 2ms/step - loss: 1.7913
Epoch 54 of 60

Testing for epoch 54 index 1:
32/32 [==============================] - 0s 980us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.2967
16/16 [==============================] - 0s 992us/step - loss: 0.8879
16/16 [==============================] - 0s 1ms/step - loss: 1.4623
16/16 [==============================] - 0s 770us/step - loss: 1.7019
16/16 [==============================] - 0s 786us/step - loss: 1.7649
16/16 [==============================] - 0s 803us/step - loss: 1.7749
16/16 [==============================] - 0s 2ms/step - loss: 1.7769
16/16 [==============================] - 0s 2ms/step - loss: 1.7777
16/16 [==============================] - 0s 982us/step - loss: 1.7784
16/16 [==============================] - 0s 2ms/step - loss: 1.7793

Testing for epoch 54 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.2942
16/16 [==============================] - 0s 925us/step - loss: 0.8801
16/16 [==============================] - 0s 846us/step - loss: 1.4515
16/16 [==============================] - 0s 787us/step - loss: 1.6891
16/16 [==============================] - 0s 2ms/step - loss: 1.7511
16/16 [==============================] - 0s 832us/step - loss: 1.7609
16/16 [==============================] - 0s 2ms/step - loss: 1.7628
16/16 [==============================] - 0s 859us/step - loss: 1.7636
16/16 [==============================] - 0s 800us/step - loss: 1.7643
16/16 [==============================] - 0s 774us/step - loss: 1.7652
Epoch 55 of 60

Testing for epoch 55 index 1:
32/32 [==============================] - 0s 898us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.2849
16/16 [==============================] - 0s 851us/step - loss: 0.9010
16/16 [==============================] - 0s 2ms/step - loss: 1.5057
16/16 [==============================] - 0s 2ms/step - loss: 1.7561
16/16 [==============================] - 0s 2ms/step - loss: 1.8207
16/16 [==============================] - 0s 2ms/step - loss: 1.8308
16/16 [==============================] - 0s 857us/step - loss: 1.8328
16/16 [==============================] - 0s 2ms/step - loss: 1.8336
16/16 [==============================] - 0s 1ms/step - loss: 1.8342
16/16 [==============================] - 0s 2ms/step - loss: 1.8350

Testing for epoch 55 index 2:
32/32 [==============================] - 0s 567us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.2748
16/16 [==============================] - 0s 802us/step - loss: 0.9142
16/16 [==============================] - 0s 2ms/step - loss: 1.5455
16/16 [==============================] - 0s 833us/step - loss: 1.8067
16/16 [==============================] - 0s 2ms/step - loss: 1.8736
16/16 [==============================] - 0s 1ms/step - loss: 1.8841
16/16 [==============================] - 0s 2ms/step - loss: 1.8861
16/16 [==============================] - 0s 2ms/step - loss: 1.8869
16/16 [==============================] - 0s 2ms/step - loss: 1.8875
16/16 [==============================] - 0s 2ms/step - loss: 1.8883
Epoch 56 of 60

Testing for epoch 56 index 1:
32/32 [==============================] - 0s 562us/step
16/16 [==============================] - 0s 810us/step - loss: 0.2756
16/16 [==============================] - 0s 1ms/step - loss: 0.9080
16/16 [==============================] - 0s 771us/step - loss: 1.5313
16/16 [==============================] - 0s 2ms/step - loss: 1.7872
16/16 [==============================] - 0s 2ms/step - loss: 1.8517
16/16 [==============================] - 0s 866us/step - loss: 1.8617
16/16 [==============================] - 0s 793us/step - loss: 1.8636
16/16 [==============================] - 0s 2ms/step - loss: 1.8644
16/16 [==============================] - 0s 2ms/step - loss: 1.8650
16/16 [==============================] - 0s 1000us/step - loss: 1.8659

Testing for epoch 56 index 2:
32/32 [==============================] - 0s 585us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.2767
16/16 [==============================] - 0s 2ms/step - loss: 0.9042
16/16 [==============================] - 0s 991us/step - loss: 1.5266
16/16 [==============================] - 0s 2ms/step - loss: 1.7809
16/16 [==============================] - 0s 781us/step - loss: 1.8447
16/16 [==============================] - 0s 1ms/step - loss: 1.8545
16/16 [==============================] - 0s 2ms/step - loss: 1.8563
16/16 [==============================] - 0s 812us/step - loss: 1.8571
16/16 [==============================] - 0s 2ms/step - loss: 1.8577
16/16 [==============================] - 0s 2ms/step - loss: 1.8586
Epoch 57 of 60

Testing for epoch 57 index 1:
32/32 [==============================] - 0s 566us/step
16/16 [==============================] - 0s 962us/step - loss: 0.2789
16/16 [==============================] - 0s 823us/step - loss: 0.9041
16/16 [==============================] - 0s 2ms/step - loss: 1.5242
16/16 [==============================] - 0s 2ms/step - loss: 1.7751
16/16 [==============================] - 0s 1ms/step - loss: 1.8372
16/16 [==============================] - 0s 2ms/step - loss: 1.8466
16/16 [==============================] - 0s 1ms/step - loss: 1.8484
16/16 [==============================] - 0s 817us/step - loss: 1.8491
16/16 [==============================] - 0s 799us/step - loss: 1.8497
16/16 [==============================] - 0s 805us/step - loss: 1.8506

Testing for epoch 57 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 789us/step - loss: 0.2721
16/16 [==============================] - 0s 836us/step - loss: 0.9176
16/16 [==============================] - 0s 1ms/step - loss: 1.5615
16/16 [==============================] - 0s 799us/step - loss: 1.8203
16/16 [==============================] - 0s 2ms/step - loss: 1.8840
16/16 [==============================] - 0s 796us/step - loss: 1.8936
16/16 [==============================] - 0s 789us/step - loss: 1.8954
16/16 [==============================] - 0s 2ms/step - loss: 1.8961
16/16 [==============================] - 0s 798us/step - loss: 1.8967
16/16 [==============================] - 0s 2ms/step - loss: 1.8976
Epoch 58 of 60

Testing for epoch 58 index 1:
32/32 [==============================] - 0s 971us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.2852
16/16 [==============================] - 0s 881us/step - loss: 0.9020
16/16 [==============================] - 0s 2ms/step - loss: 1.5168
16/16 [==============================] - 0s 922us/step - loss: 1.7593
16/16 [==============================] - 0s 828us/step - loss: 1.8184
16/16 [==============================] - 0s 771us/step - loss: 1.8271
16/16 [==============================] - 0s 815us/step - loss: 1.8288
16/16 [==============================] - 0s 958us/step - loss: 1.8295
16/16 [==============================] - 0s 787us/step - loss: 1.8301
16/16 [==============================] - 0s 788us/step - loss: 1.8310

Testing for epoch 58 index 2:
32/32 [==============================] - 0s 967us/step
16/16 [==============================] - 0s 876us/step - loss: 0.2818
16/16 [==============================] - 0s 2ms/step - loss: 0.9145
16/16 [==============================] - 0s 860us/step - loss: 1.5509
16/16 [==============================] - 0s 797us/step - loss: 1.7973
16/16 [==============================] - 0s 780us/step - loss: 1.8577
16/16 [==============================] - 0s 782us/step - loss: 1.8665
16/16 [==============================] - 0s 1ms/step - loss: 1.8682
16/16 [==============================] - 0s 951us/step - loss: 1.8689
16/16 [==============================] - 0s 764us/step - loss: 1.8695
16/16 [==============================] - 0s 777us/step - loss: 1.8704
Epoch 59 of 60

Testing for epoch 59 index 1:
32/32 [==============================] - 0s 591us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.2787
16/16 [==============================] - 0s 2ms/step - loss: 0.9060
16/16 [==============================] - 0s 2ms/step - loss: 1.5380
16/16 [==============================] - 0s 892us/step - loss: 1.7782
16/16 [==============================] - 0s 864us/step - loss: 1.8366
16/16 [==============================] - 0s 2ms/step - loss: 1.8450
16/16 [==============================] - 0s 824us/step - loss: 1.8465
16/16 [==============================] - 0s 2ms/step - loss: 1.8472
16/16 [==============================] - 0s 2ms/step - loss: 1.8478
16/16 [==============================] - 0s 1ms/step - loss: 1.8488

Testing for epoch 59 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 783us/step - loss: 0.2702
16/16 [==============================] - 0s 2ms/step - loss: 0.9158
16/16 [==============================] - 0s 930us/step - loss: 1.5716
16/16 [==============================] - 0s 2ms/step - loss: 1.8165
16/16 [==============================] - 0s 2ms/step - loss: 1.8763
16/16 [==============================] - 0s 2ms/step - loss: 1.8847
16/16 [==============================] - 0s 893us/step - loss: 1.8862
16/16 [==============================] - 0s 2ms/step - loss: 1.8869
16/16 [==============================] - 0s 2ms/step - loss: 1.8876
16/16 [==============================] - 0s 2ms/step - loss: 1.8885
Epoch 60 of 60

Testing for epoch 60 index 1:
32/32 [==============================] - 0s 564us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.2713
16/16 [==============================] - 0s 874us/step - loss: 0.9090
16/16 [==============================] - 0s 2ms/step - loss: 1.5573
16/16 [==============================] - 0s 843us/step - loss: 1.7950
16/16 [==============================] - 0s 800us/step - loss: 1.8527
16/16 [==============================] - 0s 784us/step - loss: 1.8605
16/16 [==============================] - 0s 803us/step - loss: 1.8620
16/16 [==============================] - 0s 2ms/step - loss: 1.8627
16/16 [==============================] - 0s 930us/step - loss: 1.8633
16/16 [==============================] - 0s 772us/step - loss: 1.8642

Testing for epoch 60 index 2:
32/32 [==============================] - 0s 569us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.2695
16/16 [==============================] - 0s 2ms/step - loss: 0.9206
16/16 [==============================] - 0s 830us/step - loss: 1.5858
16/16 [==============================] - 0s 880us/step - loss: 1.8277
16/16 [==============================] - 0s 820us/step - loss: 1.8862
16/16 [==============================] - 0s 820us/step - loss: 1.8940
16/16 [==============================] - 0s 784us/step - loss: 1.8955
16/16 [==============================] - 0s 2ms/step - loss: 1.8961
16/16 [==============================] - 0s 2ms/step - loss: 1.8968
16/16 [==============================] - 0s 2ms/step - loss: 1.8977
32/32 [==============================] - 0s 2ms/step

MO_GAAL(contamination=0.05, k=10, lr_d=0.01, lr_g=0.0001, momentum=0.9,
    stop_epochs=20)

outlier_MO_GAAL_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_linear,outlier_MO_GAAL_one)
_conf.conf("MO-GAAL (Liu et al., 2019)")

Accuracy: 0.947
Precision: 0.478
Recall: 0.431
F1 Score: 0.454

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.947
Precision(TP / TP + FP): 0.478
Recall(TP / TP + FN): 0.431
F1 Score(2*precision*recall/precision+recall): 0.454

fpr, tpr, thresh = roc_curve(outlier_true_linear,clf.decision_function(_df))

32/32 [==============================] - 0s 548us/step

auc(fpr, tpr)

0.5756523895121799

tab_linear = pd.concat([tab_linear,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_linear

	Accuracy	Precision	Recall	F1	AUC
GODE	0.999	1.000000	0.980392	0.990099	0.999979
LOF (Breunig et al., 2000)	0.991	0.920000	0.901961	0.910891	0.997541
kNN (Ramaswamy et al., 2000)	0.991	0.920000	0.901961	0.910891	0.997366
CBLOF (He et al., 2003)	0.969	0.700000	0.686275	0.693069	0.959214
OCSVM (Sch ̈olkopf et al., 2001)	0.923	0.370000	0.725490	0.490066	0.864150
MCD (Hardin and Rocke, 2004)	0.999	1.000000	0.980392	0.990099	0.999959
Feature Bagging (Lazarevic and Kumar, 2005)	0.993	0.940000	0.921569	0.930693	0.997397
ABOD (Kriegel et al., 2008)	0.973	0.740000	0.725490	0.732673	0.990206
Isolation Forest (Liu et al., 2008)	0.987	0.880000	0.862745	0.871287	0.995826
HBOS (Goldstein and Dengel, 2012)	0.972	0.925926	0.490196	0.641026	0.863644
SOS (Janssens et al., 2012)	0.907	0.080000	0.078431	0.079208	0.541705
SO-GAAL (Liu et al., 2019)	0.946	0.468085	0.431373	0.448980	0.575208
MO-GAAL (Liu et al., 2019)	0.947	0.478261	0.431373	0.453608	0.575652

LSCP_Linear

detectors = [KNN(), LOF(), OCSVM()]
clf = LSCP(detectors,contamination=0.05, random_state=77)
clf.fit(_df[['x', 'y']])

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/pyod/models/lscp.py:382: UserWarning: The number of histogram bins is greater than the number of classifiers, reducing n_bins to n_clf.
  warnings.warn(
/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/scipy/stats/_stats_py.py:4068: PearsonRConstantInputWarning: An input array is constant; the correlation coefficient is not defined.
  warnings.warn(PearsonRConstantInputWarning())
/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/scipy/stats/_stats_py.py:4068: PearsonRConstantInputWarning: An input array is constant; the correlation coefficient is not defined.
  warnings.warn(PearsonRConstantInputWarning())

LSCP(contamination=0.05,
   detector_list=[KNN(algorithm='auto', contamination=0.1, leaf_size=30, method='largest',
  metric='minkowski', metric_params=None, n_jobs=1, n_neighbors=5, p=2,
  radius=1.0), LOF(algorithm='auto', contamination=0.1, leaf_size=30, metric='minkowski',
  metric_params=None, n_jobs=1, n_neighbors=20, novelty=True, p=2), OCSVM(cache_size=200, coef0=0.0, contamination=0.1, degree=3, gamma='auto',
   kernel='rbf', max_iter=-1, nu=0.5, shrinking=True, tol=0.001,
   verbose=False)],
   local_max_features=1.0, local_region_size=30, n_bins=3,
   random_state=RandomState(MT19937) at 0x7FD69813D940)

outlier_LSCP_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_linear,outlier_LSCP_one)
_conf.conf("LSCP (Zhao et al., 2019)")

Accuracy: 0.985
Precision: 0.860
Recall: 0.843
F1 Score: 0.851

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.985
Precision(TP / TP + FP): 0.860
Recall(TP / TP + FN): 0.843
F1 Score(2*precision*recall/precision+recall): 0.851

fpr, tpr, thresh = roc_curve(outlier_true_linear,clf.decision_function(_df))

auc(fpr, tpr)

0.9985330275418913

tab_linear = pd.concat([tab_linear,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_linear

	Accuracy	Precision	Recall	F1	AUC
GODE	0.999	1.000000	0.980392	0.990099	0.999979
LOF (Breunig et al., 2000)	0.991	0.920000	0.901961	0.910891	0.997541
kNN (Ramaswamy et al., 2000)	0.991	0.920000	0.901961	0.910891	0.997366
CBLOF (He et al., 2003)	0.969	0.700000	0.686275	0.693069	0.959214
OCSVM (Sch ̈olkopf et al., 2001)	0.923	0.370000	0.725490	0.490066	0.864150
MCD (Hardin and Rocke, 2004)	0.999	1.000000	0.980392	0.990099	0.999959
Feature Bagging (Lazarevic and Kumar, 2005)	0.993	0.940000	0.921569	0.930693	0.997397
ABOD (Kriegel et al., 2008)	0.973	0.740000	0.725490	0.732673	0.990206
Isolation Forest (Liu et al., 2008)	0.987	0.880000	0.862745	0.871287	0.995826
HBOS (Goldstein and Dengel, 2012)	0.972	0.925926	0.490196	0.641026	0.863644
SOS (Janssens et al., 2012)	0.907	0.080000	0.078431	0.079208	0.541705
SO-GAAL (Liu et al., 2019)	0.946	0.468085	0.431373	0.448980	0.575208
MO-GAAL (Liu et al., 2019)	0.947	0.478261	0.431373	0.453608	0.575652
LSCP (Zhao et al., 2019)	0.985	0.860000	0.843137	0.851485	0.998533

tab_linear

round(tab_linear,3)

	Accuracy	Precision	Recall	F1	AUC
GODE	0.999	1.000	0.980	0.990	1.000
LOF (Breunig et al., 2000)	0.991	0.920	0.902	0.911	0.998
kNN (Ramaswamy et al., 2000)	0.991	0.920	0.902	0.911	0.997
CBLOF (He et al., 2003)	0.969	0.700	0.686	0.693	0.959
OCSVM (Sch ̈olkopf et al., 2001)	0.923	0.370	0.725	0.490	0.864
MCD (Hardin and Rocke, 2004)	0.999	1.000	0.980	0.990	1.000
Feature Bagging (Lazarevic and Kumar, 2005)	0.993	0.940	0.922	0.931	0.997
ABOD (Kriegel et al., 2008)	0.973	0.740	0.725	0.733	0.990
Isolation Forest (Liu et al., 2008)	0.987	0.880	0.863	0.871	0.996
HBOS (Goldstein and Dengel, 2012)	0.972	0.926	0.490	0.641	0.864
SOS (Janssens et al., 2012)	0.907	0.080	0.078	0.079	0.542
SO-GAAL (Liu et al., 2019)	0.946	0.468	0.431	0.449	0.575
MO-GAAL (Liu et al., 2019)	0.947	0.478	0.431	0.454	0.576
LSCP (Zhao et al., 2019)	0.985	0.860	0.843	0.851	0.999

Orbit

# np.random.seed(1212)
# epsilon = np.around(np.random.normal(size=1000),15)
# signal = np.random.choice(np.concatenate((np.random.uniform(-7, -5, 25).round(15), np.random.uniform(5, 7, 25).round(15), np.repeat(0, 950))), 1000)
# eta = signal + epsilon
# np.random.seed(777)
# pi=np.pi
# n=1000
# ang=np.linspace(-pi,pi-2*pi/n,n)
# r=5+np.cos(np.linspace(0,12*pi,n))
# vx=r*np.cos(ang)
# vy=r*np.sin(ang)
# f1=10*np.sin(np.linspace(0,6*pi,n))
# f = f1 + eta
# _df = pd.DataFrame({'x' : vx, 'y' : vy, 'f' : f})
# outlier_true_orbit = signal.copy()
# outlier_true_orbit = list(map(lambda x: 1 if x!=0 else 0,outlier_true_orbit))

tab_orbit = pd.DataFrame(columns=["Accuracy","Precision","Recall","F1","AUC"])

np.random.seed(777)
epsilon = np.around(np.random.normal(size=1000),15)
signal = np.random.choice(np.concatenate((np.random.uniform(-4, -1, 25).round(15), np.random.uniform(1, 4, 25).round(15), np.repeat(0, 950))), 1000)
eta = signal + epsilon
pi=np.pi
n=1000
ang=np.linspace(-pi,pi-2*pi/n,n)
r=5+np.cos(np.linspace(0,12*pi,n))
vx=r*np.cos(ang)
vy=r*np.sin(ang)
f1=10*np.sin(np.linspace(0,6*pi,n))
f = f1 + eta
_df = pd.DataFrame({'x' : vx, 'y' : vy, 'f' : f})
outlier_true_orbit = signal.copy()
outlier_true_orbit = list(map(lambda x: 1 if x!=0 else 0,outlier_true_orbit))

GODE_Orbit

_Orbit = Orbit(_df)
_Orbit.get_distance()

100%|██████████| 1000/1000 [00:01<00:00, 595.48it/s]

_Orbit.get_weightmatrix(theta=(_Orbit.D[_Orbit.D>0].mean()),kappa=2500) 
_Orbit.fit(sd=15,ref=20)

outlier_GODE_one_old = (_Orbit.df['Residual']**2).tolist()
sorted_data = sorted(outlier_GODE_one_old,reverse=True)
index = int(len(sorted_data) * 0.05)
five_percent = sorted_data[index]
outlier_GODE_one = list(map(lambda x: 1 if x > five_percent else 0,outlier_GODE_one_old))

_conf = Conf_matrx(outlier_true_orbit,outlier_GODE_one)
_conf.conf("GODE")

Accuracy: 0.961
Precision: 0.600
Recall: 0.612
F1 Score: 0.606

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.961
Precision(TP / TP + FP): 0.600
Recall(TP / TP + FN): 0.612
F1 Score(2*precision*recall/precision+recall): 0.606

fpr, tpr, thresh = roc_curve(outlier_true_orbit,outlier_GODE_one_old)

auc(fpr, tpr)

0.893023455438958

tab_orbit = pd.concat([tab_orbit,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_orbit

	Accuracy	Precision	Recall	F1	AUC
GODE	0.961	0.6	0.612245	0.606061	0.893023

LOF_Orbit

np.random.seed(77)
clf = LOF(contamination=0.05)
clf.fit(_df[['x', 'y','f']])

LOF(algorithm='auto', contamination=0.05, leaf_size=30, metric='minkowski',
  metric_params=None, n_jobs=1, n_neighbors=20, novelty=True, p=2)

outlier_LOF_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_orbit,clf.fit_predict(_df))
_conf.conf("LOF (Breunig et al., 2000)")

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/utils/deprecation.py:86: FutureWarning: Function fit_predict is deprecated
  warnings.warn(msg, category=FutureWarning)

Accuracy: 0.921
Precision: 0.200
Recall: 0.204
F1 Score: 0.202

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.921
Precision(TP / TP + FP): 0.200
Recall(TP / TP + FN): 0.204
F1 Score(2*precision*recall/precision+recall): 0.202

fpr, tpr, thresh = roc_curve(outlier_true_orbit,clf.decision_function(_df))

auc(fpr, tpr)

0.6641236936414945

tab_orbit = pd.concat([tab_orbit,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_orbit

	Accuracy	Precision	Recall	F1	AUC
GODE	0.961	0.6	0.612245	0.606061	0.893023
LOF (Breunig et al., 2000)	0.921	0.2	0.204082	0.202020	0.664124

KNN_Orbit

np.random.seed(77)
clf = KNN(contamination=0.05)
clf.fit(_df[['x', 'y','f']])

KNN(algorithm='auto', contamination=0.05, leaf_size=30, method='largest',
  metric='minkowski', metric_params=None, n_jobs=1, n_neighbors=5, p=2,
  radius=1.0)

outlier_KNN_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_orbit,outlier_KNN_one)
_conf.conf("kNN (Ramaswamy et al., 2000)")

Accuracy: 0.947
Precision: 0.460
Recall: 0.469
F1 Score: 0.465

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.947
Precision(TP / TP + FP): 0.460
Recall(TP / TP + FN): 0.469
F1 Score(2*precision*recall/precision+recall): 0.465

fpr, tpr, thresh = roc_curve(outlier_true_orbit,clf.decision_function(_df))

auc(fpr, tpr)

0.8479903002210347

tab_orbit = pd.concat([tab_orbit,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_orbit

	Accuracy	Precision	Recall	F1	AUC
GODE	0.961	0.60	0.612245	0.606061	0.893023
LOF (Breunig et al., 2000)	0.921	0.20	0.204082	0.202020	0.664124
kNN (Ramaswamy et al., 2000)	0.947	0.46	0.469388	0.464646	0.847990

CBLOF_Orbit

clf = CBLOF(contamination=0.05,random_state=77)
clf.fit(_df[['x', 'y','f']])

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning
  super()._check_params_vs_input(X, default_n_init=10)

CBLOF(alpha=0.9, beta=5, check_estimator=False, clustering_estimator=None,
   contamination=0.05, n_clusters=8, n_jobs=None, random_state=77,
   use_weights=False)

outlier_CBLOF_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_orbit,outlier_CBLOF_one)
_conf.conf("CBLOF (He et al., 2003)")

Accuracy: 0.911
Precision: 0.100
Recall: 0.102
F1 Score: 0.101

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.911
Precision(TP / TP + FP): 0.100
Recall(TP / TP + FN): 0.102
F1 Score(2*precision*recall/precision+recall): 0.101

fpr, tpr, thresh = roc_curve(outlier_true_orbit,clf.decision_function(_df))

auc(fpr, tpr)

0.533402004334857

tab_orbit = pd.concat([tab_orbit,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_orbit

	Accuracy	Precision	Recall	F1	AUC
GODE	0.961	0.60	0.612245	0.606061	0.893023
LOF (Breunig et al., 2000)	0.921	0.20	0.204082	0.202020	0.664124
kNN (Ramaswamy et al., 2000)	0.947	0.46	0.469388	0.464646	0.847990
CBLOF (He et al., 2003)	0.911	0.10	0.102041	0.101010	0.533402

OCSVM_Orbit

np.random.seed(77)
clf = OCSVM(nu=0.05)
clf.fit(_df)

OCSVM(cache_size=200, coef0=0.0, contamination=0.1, degree=3, gamma='auto',
   kernel='rbf', max_iter=-1, nu=0.05, shrinking=True, tol=0.001,
   verbose=False)

outlier_OSVM_one = list(clf.predict(_df))

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/base.py:457: UserWarning: X has feature names, but OneClassSVM was fitted without feature names
  warnings.warn(

_conf = Conf_matrx(outlier_true_orbit,outlier_OSVM_one)
_conf.conf("OCSVM (Sch ̈olkopf et al., 2001)")

Accuracy: 0.893
Precision: 0.210
Recall: 0.429
F1 Score: 0.282

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.893
Precision(TP / TP + FP): 0.210
Recall(TP / TP + FN): 0.429
F1 Score(2*precision*recall/precision+recall): 0.282

fpr, tpr, thresh = roc_curve(outlier_true_orbit,clf.decision_function(_df))

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/base.py:457: UserWarning: X has feature names, but OneClassSVM was fitted without feature names
  warnings.warn(

auc(fpr, tpr)

0.7887508315629091

tab_orbit = pd.concat([tab_orbit,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_orbit

	Accuracy	Precision	Recall	F1	AUC
GODE	0.961	0.60	0.612245	0.606061	0.893023
LOF (Breunig et al., 2000)	0.921	0.20	0.204082	0.202020	0.664124
kNN (Ramaswamy et al., 2000)	0.947	0.46	0.469388	0.464646	0.847990
CBLOF (He et al., 2003)	0.911	0.10	0.102041	0.101010	0.533402
OCSVM (Sch ̈olkopf et al., 2001)	0.893	0.21	0.428571	0.281879	0.788751

MCD_Orbit

clf = MCD(contamination=0.05 , random_state = 77)
clf.fit(_df[['x','y','f']])

MCD(assume_centered=False, contamination=0.05, random_state=77,
  store_precision=True, support_fraction=None)

outlier_MCD_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_orbit,outlier_MCD_one)
_conf.conf("MCD (Hardin and Rocke, 2004)")

Accuracy: 0.911
Precision: 0.100
Recall: 0.102
F1 Score: 0.101

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.911
Precision(TP / TP + FP): 0.100
Recall(TP / TP + FN): 0.102
F1 Score(2*precision*recall/precision+recall): 0.101

fpr, tpr, thresh = roc_curve(outlier_true_orbit,clf.decision_function(_df))

auc(fpr, tpr)

0.45400115882315073

tab_orbit = pd.concat([tab_orbit,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_orbit

	Accuracy	Precision	Recall	F1	AUC
GODE	0.961	0.60	0.612245	0.606061	0.893023
LOF (Breunig et al., 2000)	0.921	0.20	0.204082	0.202020	0.664124
kNN (Ramaswamy et al., 2000)	0.947	0.46	0.469388	0.464646	0.847990
CBLOF (He et al., 2003)	0.911	0.10	0.102041	0.101010	0.533402
OCSVM (Sch ̈olkopf et al., 2001)	0.893	0.21	0.428571	0.281879	0.788751
MCD (Hardin and Rocke, 2004)	0.911	0.10	0.102041	0.101010	0.454001

Feature Bagging_Orbit

clf = FeatureBagging(contamination=0.05, random_state=77)
clf.fit(_df[['x', 'y','f']])

FeatureBagging(base_estimator=None, bootstrap_features=False,
        check_detector=True, check_estimator=False, combination='average',
        contamination=0.05, estimator_params={}, max_features=1.0,
        n_estimators=10, n_jobs=1, random_state=77, verbose=0)

outlier_FeatureBagging_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_orbit,outlier_FeatureBagging_one)
_conf.conf("Feature Bagging (Lazarevic and Kumar, 2005)")

Accuracy: 0.921
Precision: 0.200
Recall: 0.204
F1 Score: 0.202

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.921
Precision(TP / TP + FP): 0.200
Recall(TP / TP + FN): 0.204
F1 Score(2*precision*recall/precision+recall): 0.202

fpr, tpr, thresh = roc_curve(outlier_true_orbit,clf.decision_function(_df))

auc(fpr, tpr)

0.6779973819180669

tab_orbit = pd.concat([tab_orbit,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_orbit

	Accuracy	Precision	Recall	F1	AUC
GODE	0.961	0.60	0.612245	0.606061	0.893023
LOF (Breunig et al., 2000)	0.921	0.20	0.204082	0.202020	0.664124
kNN (Ramaswamy et al., 2000)	0.947	0.46	0.469388	0.464646	0.847990
CBLOF (He et al., 2003)	0.911	0.10	0.102041	0.101010	0.533402
OCSVM (Sch ̈olkopf et al., 2001)	0.893	0.21	0.428571	0.281879	0.788751
MCD (Hardin and Rocke, 2004)	0.911	0.10	0.102041	0.101010	0.454001
Feature Bagging (Lazarevic and Kumar, 2005)	0.921	0.20	0.204082	0.202020	0.677997

ABOD_Orbit

np.random.seed(77)
clf = ABOD(contamination=0.05)
clf.fit(_df[['x', 'y','f']])

ABOD(contamination=0.05, method='fast', n_neighbors=5)

outlier_ABOD_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_orbit,outlier_ABOD_one)
_conf.conf("ABOD (Kriegel et al., 2008)")

Accuracy: 0.951
Precision: 0.500
Recall: 0.510
F1 Score: 0.505

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.951
Precision(TP / TP + FP): 0.500
Recall(TP / TP + FN): 0.510
F1 Score(2*precision*recall/precision+recall): 0.505

fpr, tpr, thresh = roc_curve(outlier_true_orbit,clf.decision_function(_df))

auc(fpr, tpr)

0.8639241185433164

tab_orbit = pd.concat([tab_orbit,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_orbit

	Accuracy	Precision	Recall	F1	AUC
GODE	0.961	0.60	0.612245	0.606061	0.893023
LOF (Breunig et al., 2000)	0.921	0.20	0.204082	0.202020	0.664124
kNN (Ramaswamy et al., 2000)	0.947	0.46	0.469388	0.464646	0.847990
CBLOF (He et al., 2003)	0.911	0.10	0.102041	0.101010	0.533402
OCSVM (Sch ̈olkopf et al., 2001)	0.893	0.21	0.428571	0.281879	0.788751
MCD (Hardin and Rocke, 2004)	0.911	0.10	0.102041	0.101010	0.454001
Feature Bagging (Lazarevic and Kumar, 2005)	0.921	0.20	0.204082	0.202020	0.677997
ABOD (Kriegel et al., 2008)	0.951	0.50	0.510204	0.505051	0.863924

IForest_Orbit

clf = IForest(contamination=0.05,random_state=77)
clf.fit(_df[['x', 'y','f']])

IForest(behaviour='old', bootstrap=False, contamination=0.05,
    max_features=1.0, max_samples='auto', n_estimators=100, n_jobs=1,
    random_state=77, verbose=0)

outlier_IForest_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_orbit,outlier_IForest_one)
_conf.conf("Isolation Forest (Liu et al., 2008)")

Accuracy: 0.925
Precision: 0.240
Recall: 0.245
F1 Score: 0.242

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.925
Precision(TP / TP + FP): 0.240
Recall(TP / TP + FN): 0.245
F1 Score(2*precision*recall/precision+recall): 0.242

fpr, tpr, thresh = roc_curve(outlier_true_orbit,clf.decision_function(_df))

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/base.py:457: UserWarning: X has feature names, but IsolationForest was fitted without feature names
  warnings.warn(

auc(fpr, tpr)

0.6180175540247645

tab_orbit = pd.concat([tab_orbit,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_orbit

	Accuracy	Precision	Recall	F1	AUC
GODE	0.961	0.60	0.612245	0.606061	0.893023
LOF (Breunig et al., 2000)	0.921	0.20	0.204082	0.202020	0.664124
kNN (Ramaswamy et al., 2000)	0.947	0.46	0.469388	0.464646	0.847990
CBLOF (He et al., 2003)	0.911	0.10	0.102041	0.101010	0.533402
OCSVM (Sch ̈olkopf et al., 2001)	0.893	0.21	0.428571	0.281879	0.788751
MCD (Hardin and Rocke, 2004)	0.911	0.10	0.102041	0.101010	0.454001
Feature Bagging (Lazarevic and Kumar, 2005)	0.921	0.20	0.204082	0.202020	0.677997
ABOD (Kriegel et al., 2008)	0.951	0.50	0.510204	0.505051	0.863924
Isolation Forest (Liu et al., 2008)	0.925	0.24	0.244898	0.242424	0.618018

HBOS_Orbit

np.random.seed(77)
clf = HBOS(contamination=0.05)
clf.fit(_df[['x', 'y','f']])

HBOS(alpha=0.1, contamination=0.05, n_bins=10, tol=0.5)

outlier_HBOS_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_orbit,outlier_HBOS_one)
_conf.conf("HBOS (Goldstein and Dengel, 2012)")

Accuracy: 0.921
Precision: 0.105
Recall: 0.082
F1 Score: 0.092

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.921
Precision(TP / TP + FP): 0.105
Recall(TP / TP + FN): 0.082
F1 Score(2*precision*recall/precision+recall): 0.092

fpr, tpr, thresh = roc_curve(outlier_true_orbit,clf.decision_function(_df))

auc(fpr, tpr)

0.5297431275349256

tab_orbit = pd.concat([tab_orbit,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_orbit

	Accuracy	Precision	Recall	F1	AUC
GODE	0.961	0.600000	0.612245	0.606061	0.893023
LOF (Breunig et al., 2000)	0.921	0.200000	0.204082	0.202020	0.664124
kNN (Ramaswamy et al., 2000)	0.947	0.460000	0.469388	0.464646	0.847990
CBLOF (He et al., 2003)	0.911	0.100000	0.102041	0.101010	0.533402
OCSVM (Sch ̈olkopf et al., 2001)	0.893	0.210000	0.428571	0.281879	0.788751
MCD (Hardin and Rocke, 2004)	0.911	0.100000	0.102041	0.101010	0.454001
Feature Bagging (Lazarevic and Kumar, 2005)	0.921	0.200000	0.204082	0.202020	0.677997
ABOD (Kriegel et al., 2008)	0.951	0.500000	0.510204	0.505051	0.863924
Isolation Forest (Liu et al., 2008)	0.925	0.240000	0.244898	0.242424	0.618018
HBOS (Goldstein and Dengel, 2012)	0.921	0.105263	0.081633	0.091954	0.529743

SOS_Orbit

np.random.seed(77)
clf = SOS(contamination=0.05)
clf.fit(_df[['x', 'y','f']])

SOS(contamination=0.05, eps=1e-05, metric='euclidean', perplexity=4.5)

outlier_SOS_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_orbit,outlier_SOS_one)
_conf.conf("SOS (Janssens et al., 2012)")

Accuracy: 0.941
Precision: 0.400
Recall: 0.408
F1 Score: 0.404

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.941
Precision(TP / TP + FP): 0.400
Recall(TP / TP + FN): 0.408
F1 Score(2*precision*recall/precision+recall): 0.404

fpr, tpr, thresh = roc_curve(outlier_true_orbit,clf.decision_function(_df))

auc(fpr, tpr)

0.8442241249812228

tab_orbit = pd.concat([tab_orbit,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_orbit

	Accuracy	Precision	Recall	F1	AUC
GODE	0.961	0.600000	0.612245	0.606061	0.893023
LOF (Breunig et al., 2000)	0.921	0.200000	0.204082	0.202020	0.664124
kNN (Ramaswamy et al., 2000)	0.947	0.460000	0.469388	0.464646	0.847990
CBLOF (He et al., 2003)	0.911	0.100000	0.102041	0.101010	0.533402
OCSVM (Sch ̈olkopf et al., 2001)	0.893	0.210000	0.428571	0.281879	0.788751
MCD (Hardin and Rocke, 2004)	0.911	0.100000	0.102041	0.101010	0.454001
Feature Bagging (Lazarevic and Kumar, 2005)	0.921	0.200000	0.204082	0.202020	0.677997
ABOD (Kriegel et al., 2008)	0.951	0.500000	0.510204	0.505051	0.863924
Isolation Forest (Liu et al., 2008)	0.925	0.240000	0.244898	0.242424	0.618018
HBOS (Goldstein and Dengel, 2012)	0.921	0.105263	0.081633	0.091954	0.529743
SOS (Janssens et al., 2012)	0.941	0.400000	0.408163	0.404040	0.844224

SO_GAAL_Orbit

np.random.seed(77)
clf = SO_GAAL(contamination=0.05)
clf.fit(_df[['x', 'y','f']])

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/keras/src/optimizers/legacy/gradient_descent.py:114: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.
  super().__init__(name, **kwargs)

Epoch 1 of 60

Testing for epoch 1 index 1:

Testing for epoch 1 index 2:
Epoch 2 of 60

Testing for epoch 2 index 1:

Testing for epoch 2 index 2:
Epoch 3 of 60

Testing for epoch 3 index 1:

Testing for epoch 3 index 2:
Epoch 4 of 60

Testing for epoch 4 index 1:

Testing for epoch 4 index 2:
Epoch 5 of 60

Testing for epoch 5 index 1:

Testing for epoch 5 index 2:
Epoch 6 of 60

Testing for epoch 6 index 1:

Testing for epoch 6 index 2:
Epoch 7 of 60

Testing for epoch 7 index 1:

Testing for epoch 7 index 2:
Epoch 8 of 60

Testing for epoch 8 index 1:

Testing for epoch 8 index 2:
Epoch 9 of 60

Testing for epoch 9 index 1:

Testing for epoch 9 index 2:
Epoch 10 of 60

Testing for epoch 10 index 1:

Testing for epoch 10 index 2:
Epoch 11 of 60

Testing for epoch 11 index 1:

Testing for epoch 11 index 2:
Epoch 12 of 60

Testing for epoch 12 index 1:

Testing for epoch 12 index 2:
Epoch 13 of 60

Testing for epoch 13 index 1:

Testing for epoch 13 index 2:
Epoch 14 of 60

Testing for epoch 14 index 1:

Testing for epoch 14 index 2:
Epoch 15 of 60

Testing for epoch 15 index 1:

Testing for epoch 15 index 2:
Epoch 16 of 60

Testing for epoch 16 index 1:

Testing for epoch 16 index 2:
Epoch 17 of 60

Testing for epoch 17 index 1:

Testing for epoch 17 index 2:
Epoch 18 of 60

Testing for epoch 18 index 1:

Testing for epoch 18 index 2:
Epoch 19 of 60

Testing for epoch 19 index 1:

Testing for epoch 19 index 2:
Epoch 20 of 60

Testing for epoch 20 index 1:

Testing for epoch 20 index 2:
Epoch 21 of 60

Testing for epoch 21 index 1:

Testing for epoch 21 index 2:
Epoch 22 of 60

Testing for epoch 22 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.2682

Testing for epoch 22 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.2809
Epoch 23 of 60

Testing for epoch 23 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.2533

Testing for epoch 23 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.2402
Epoch 24 of 60

Testing for epoch 24 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.2311

Testing for epoch 24 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.2195
Epoch 25 of 60

Testing for epoch 25 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.2273

Testing for epoch 25 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.2363
Epoch 26 of 60

Testing for epoch 26 index 1:
16/16 [==============================] - 0s 825us/step - loss: 1.2320

Testing for epoch 26 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.2487
Epoch 27 of 60

Testing for epoch 27 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.2435

Testing for epoch 27 index 2:
16/16 [==============================] - 0s 826us/step - loss: 1.2753
Epoch 28 of 60

Testing for epoch 28 index 1:
16/16 [==============================] - 0s 781us/step - loss: 1.2691

Testing for epoch 28 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 1.2997
Epoch 29 of 60

Testing for epoch 29 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.3105

Testing for epoch 29 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.3315
Epoch 30 of 60

Testing for epoch 30 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 1.3413

Testing for epoch 30 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 1.3642
Epoch 31 of 60

Testing for epoch 31 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 1.3953

Testing for epoch 31 index 2:
16/16 [==============================] - 0s 790us/step - loss: 1.4018
Epoch 32 of 60

Testing for epoch 32 index 1:
16/16 [==============================] - 0s 802us/step - loss: 1.4309

Testing for epoch 32 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 1.4460
Epoch 33 of 60

Testing for epoch 33 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.4762

Testing for epoch 33 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.5035
Epoch 34 of 60

Testing for epoch 34 index 1:
16/16 [==============================] - 0s 834us/step - loss: 1.5015

Testing for epoch 34 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.5210
Epoch 35 of 60

Testing for epoch 35 index 1:
16/16 [==============================] - 0s 937us/step - loss: 1.5350

Testing for epoch 35 index 2:
16/16 [==============================] - 0s 789us/step - loss: 1.5497
Epoch 36 of 60

Testing for epoch 36 index 1:
16/16 [==============================] - 0s 790us/step - loss: 1.5742

Testing for epoch 36 index 2:
16/16 [==============================] - 0s 842us/step - loss: 1.5749
Epoch 37 of 60

Testing for epoch 37 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.5844

Testing for epoch 37 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 1.5945
Epoch 38 of 60

Testing for epoch 38 index 1:
16/16 [==============================] - 0s 926us/step - loss: 1.6030

Testing for epoch 38 index 2:
16/16 [==============================] - 0s 866us/step - loss: 1.6226
Epoch 39 of 60

Testing for epoch 39 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 1.6200

Testing for epoch 39 index 2:
16/16 [==============================] - 0s 801us/step - loss: 1.6314
Epoch 40 of 60

Testing for epoch 40 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.6379

Testing for epoch 40 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.6526
Epoch 41 of 60

Testing for epoch 41 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 1.6465

Testing for epoch 41 index 2:
16/16 [==============================] - 0s 811us/step - loss: 1.6509
Epoch 42 of 60

Testing for epoch 42 index 1:
16/16 [==============================] - 0s 800us/step - loss: 1.6508

Testing for epoch 42 index 2:
16/16 [==============================] - 0s 799us/step - loss: 1.6638
Epoch 43 of 60

Testing for epoch 43 index 1:
16/16 [==============================] - 0s 953us/step - loss: 1.6715

Testing for epoch 43 index 2:
16/16 [==============================] - 0s 780us/step - loss: 1.6817
Epoch 44 of 60

Testing for epoch 44 index 1:
16/16 [==============================] - 0s 799us/step - loss: 1.6805

Testing for epoch 44 index 2:
16/16 [==============================] - 0s 783us/step - loss: 1.6850
Epoch 45 of 60

Testing for epoch 45 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 1.7113

Testing for epoch 45 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.7162
Epoch 46 of 60

Testing for epoch 46 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.7183

Testing for epoch 46 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 1.7208
Epoch 47 of 60

Testing for epoch 47 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.7249

Testing for epoch 47 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 1.7667
Epoch 48 of 60

Testing for epoch 48 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 1.7558

Testing for epoch 48 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.7659
Epoch 49 of 60

Testing for epoch 49 index 1:
16/16 [==============================] - 0s 863us/step - loss: 1.7905

Testing for epoch 49 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.8042
Epoch 50 of 60

Testing for epoch 50 index 1:
16/16 [==============================] - 0s 818us/step - loss: 1.8058

Testing for epoch 50 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.8257
Epoch 51 of 60

Testing for epoch 51 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.8315

Testing for epoch 51 index 2:
16/16 [==============================] - 0s 930us/step - loss: 1.8521
Epoch 52 of 60

Testing for epoch 52 index 1:
16/16 [==============================] - 0s 814us/step - loss: 1.8571

Testing for epoch 52 index 2:
16/16 [==============================] - 0s 815us/step - loss: 1.8628
Epoch 53 of 60

Testing for epoch 53 index 1:
16/16 [==============================] - 0s 829us/step - loss: 1.8641

Testing for epoch 53 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.8893
Epoch 54 of 60

Testing for epoch 54 index 1:
16/16 [==============================] - 0s 896us/step - loss: 1.8914

Testing for epoch 54 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.9188
Epoch 55 of 60

Testing for epoch 55 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.9080

Testing for epoch 55 index 2:
16/16 [==============================] - 0s 820us/step - loss: 1.9341
Epoch 56 of 60

Testing for epoch 56 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 1.9327

Testing for epoch 56 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.9627
Epoch 57 of 60

Testing for epoch 57 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.9587

Testing for epoch 57 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.9495
Epoch 58 of 60

Testing for epoch 58 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 1.9537

Testing for epoch 58 index 2:
16/16 [==============================] - 0s 878us/step - loss: 1.9898
Epoch 59 of 60

Testing for epoch 59 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.9983

Testing for epoch 59 index 2:
16/16 [==============================] - 0s 960us/step - loss: 2.0003
Epoch 60 of 60

Testing for epoch 60 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.0136

Testing for epoch 60 index 2:
16/16 [==============================] - 0s 811us/step - loss: 2.0044
32/32 [==============================] - 0s 1ms/step

SO_GAAL(contamination=0.05, lr_d=0.01, lr_g=0.0001, momentum=0.9,
    stop_epochs=20)

outlier_SO_GAAL_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_orbit,outlier_SO_GAAL_one)
_conf.conf("SO-GAAL (Liu et al., 2019)")

Accuracy: 0.951
Precision: 0.000
Recall: 0.000
F1 Score: 0.000

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/metrics/_classification.py:1469: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.
  _warn_prf(average, modifier, msg_start, len(result))

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.951
Precision(TP / TP + FP): nan
Recall(TP / TP + FN): 0.000
F1 Score(2*precision*recall/precision+recall): nan

/tmp/ipykernel_3852735/4166638268.py:3: RuntimeWarning: invalid value encountered in long_scalars
  print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
/tmp/ipykernel_3852735/4166638268.py:5: RuntimeWarning: invalid value encountered in long_scalars
  print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

fpr, tpr, thresh = roc_curve(outlier_true_orbit,clf.decision_function(_df))

32/32 [==============================] - 0s 568us/step

auc(fpr, tpr)

0.46434472842764857

tab_orbit = pd.concat([tab_orbit,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_orbit

	Accuracy	Precision	Recall	F1	AUC
GODE	0.961	0.600000	0.612245	0.606061	0.893023
LOF (Breunig et al., 2000)	0.921	0.200000	0.204082	0.202020	0.664124
kNN (Ramaswamy et al., 2000)	0.947	0.460000	0.469388	0.464646	0.847990
CBLOF (He et al., 2003)	0.911	0.100000	0.102041	0.101010	0.533402
OCSVM (Sch ̈olkopf et al., 2001)	0.893	0.210000	0.428571	0.281879	0.788751
MCD (Hardin and Rocke, 2004)	0.911	0.100000	0.102041	0.101010	0.454001
Feature Bagging (Lazarevic and Kumar, 2005)	0.921	0.200000	0.204082	0.202020	0.677997
ABOD (Kriegel et al., 2008)	0.951	0.500000	0.510204	0.505051	0.863924
Isolation Forest (Liu et al., 2008)	0.925	0.240000	0.244898	0.242424	0.618018
HBOS (Goldstein and Dengel, 2012)	0.921	0.105263	0.081633	0.091954	0.529743
SOS (Janssens et al., 2012)	0.941	0.400000	0.408163	0.404040	0.844224
SO-GAAL (Liu et al., 2019)	0.951	0.000000	0.000000	0.000000	0.464345

MO_GAAL_Orbit

np.random.seed(77)
clf = MO_GAAL(contamination=0.05)
clf.fit(_df[['x', 'y','f']])

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/keras/src/optimizers/legacy/gradient_descent.py:114: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.
  super().__init__(name, **kwargs)

Epoch 1 of 60

Testing for epoch 1 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 1 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 2 of 60

Testing for epoch 2 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 2 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 3 of 60

Testing for epoch 3 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 3 index 2:
32/32 [==============================] - 0s 581us/step
Epoch 4 of 60

Testing for epoch 4 index 1:
32/32 [==============================] - 0s 583us/step

Testing for epoch 4 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 5 of 60

Testing for epoch 5 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 5 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 6 of 60

Testing for epoch 6 index 1:
32/32 [==============================] - 0s 561us/step

Testing for epoch 6 index 2:
32/32 [==============================] - 0s 601us/step
Epoch 7 of 60

Testing for epoch 7 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 7 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 8 of 60

Testing for epoch 8 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 8 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 9 of 60

Testing for epoch 9 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 9 index 2:
32/32 [==============================] - 0s 807us/step
Epoch 10 of 60

Testing for epoch 10 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 10 index 2:
32/32 [==============================] - 0s 575us/step
Epoch 11 of 60

Testing for epoch 11 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 11 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 12 of 60

Testing for epoch 12 index 1:
32/32 [==============================] - 0s 564us/step

Testing for epoch 12 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 13 of 60

Testing for epoch 13 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 13 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 14 of 60

Testing for epoch 14 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 14 index 2:
32/32 [==============================] - 0s 575us/step
Epoch 15 of 60

Testing for epoch 15 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 15 index 2:
32/32 [==============================] - 0s 565us/step
Epoch 16 of 60

Testing for epoch 16 index 1:
32/32 [==============================] - 0s 562us/step

Testing for epoch 16 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 17 of 60

Testing for epoch 17 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 17 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 18 of 60

Testing for epoch 18 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 18 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 19 of 60

Testing for epoch 19 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 19 index 2:
32/32 [==============================] - 0s 1ms/step
Epoch 20 of 60

Testing for epoch 20 index 1:
32/32 [==============================] - 0s 581us/step

Testing for epoch 20 index 2:
32/32 [==============================] - 0s 595us/step
Epoch 21 of 60

Testing for epoch 21 index 1:
32/32 [==============================] - 0s 1ms/step

Testing for epoch 21 index 2:
32/32 [==============================] - 0s 615us/step
16/16 [==============================] - 0s 976us/step - loss: 0.6211
16/16 [==============================] - 0s 864us/step - loss: 1.0813
16/16 [==============================] - 0s 847us/step - loss: 1.1397
16/16 [==============================] - 0s 1ms/step - loss: 1.1436
16/16 [==============================] - 0s 1ms/step - loss: 1.1446
16/16 [==============================] - 0s 979us/step - loss: 1.1448
16/16 [==============================] - 0s 2ms/step - loss: 1.1448
16/16 [==============================] - 0s 2ms/step - loss: 1.1447
16/16 [==============================] - 0s 2ms/step - loss: 1.1446
16/16 [==============================] - 0s 2ms/step - loss: 1.1446
Epoch 22 of 60

Testing for epoch 22 index 1:
32/32 [==============================] - 0s 888us/step
16/16 [==============================] - 0s 936us/step - loss: 0.6128
16/16 [==============================] - 0s 2ms/step - loss: 1.1006
16/16 [==============================] - 0s 2ms/step - loss: 1.1618
16/16 [==============================] - 0s 967us/step - loss: 1.1658
16/16 [==============================] - 0s 865us/step - loss: 1.1669
16/16 [==============================] - 0s 2ms/step - loss: 1.1671
16/16 [==============================] - 0s 1ms/step - loss: 1.1671
16/16 [==============================] - 0s 1ms/step - loss: 1.1670
16/16 [==============================] - 0s 2ms/step - loss: 1.1669
16/16 [==============================] - 0s 2ms/step - loss: 1.1669

Testing for epoch 22 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.6131
16/16 [==============================] - 0s 907us/step - loss: 1.1060
16/16 [==============================] - 0s 947us/step - loss: 1.1673
16/16 [==============================] - 0s 2ms/step - loss: 1.1713
16/16 [==============================] - 0s 916us/step - loss: 1.1723
16/16 [==============================] - 0s 983us/step - loss: 1.1725
16/16 [==============================] - 0s 2ms/step - loss: 1.1725
16/16 [==============================] - 0s 886us/step - loss: 1.1724
16/16 [==============================] - 0s 861us/step - loss: 1.1723
16/16 [==============================] - 0s 841us/step - loss: 1.1723
Epoch 23 of 60

Testing for epoch 23 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.6029
16/16 [==============================] - 0s 2ms/step - loss: 1.1322
16/16 [==============================] - 0s 897us/step - loss: 1.2002
16/16 [==============================] - 0s 1ms/step - loss: 1.2044
16/16 [==============================] - 0s 2ms/step - loss: 1.2055
16/16 [==============================] - 0s 1ms/step - loss: 1.2057
16/16 [==============================] - 0s 2ms/step - loss: 1.2057
16/16 [==============================] - 0s 2ms/step - loss: 1.2056
16/16 [==============================] - 0s 948us/step - loss: 1.2055
16/16 [==============================] - 0s 892us/step - loss: 1.2055

Testing for epoch 23 index 2:
32/32 [==============================] - 0s 826us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.6025
16/16 [==============================] - 0s 869us/step - loss: 1.1440
16/16 [==============================] - 0s 2ms/step - loss: 1.2117
16/16 [==============================] - 0s 854us/step - loss: 1.2159
16/16 [==============================] - 0s 2ms/step - loss: 1.2169
16/16 [==============================] - 0s 875us/step - loss: 1.2171
16/16 [==============================] - 0s 1ms/step - loss: 1.2171
16/16 [==============================] - 0s 2ms/step - loss: 1.2170
16/16 [==============================] - 0s 863us/step - loss: 1.2170
16/16 [==============================] - 0s 867us/step - loss: 1.2170
Epoch 24 of 60

Testing for epoch 24 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.5927
16/16 [==============================] - 0s 2ms/step - loss: 1.1665
16/16 [==============================] - 0s 998us/step - loss: 1.2374
16/16 [==============================] - 0s 823us/step - loss: 1.2419
16/16 [==============================] - 0s 934us/step - loss: 1.2430
16/16 [==============================] - 0s 829us/step - loss: 1.2432
16/16 [==============================] - 0s 2ms/step - loss: 1.2432
16/16 [==============================] - 0s 2ms/step - loss: 1.2431
16/16 [==============================] - 0s 2ms/step - loss: 1.2431
16/16 [==============================] - 0s 2ms/step - loss: 1.2431

Testing for epoch 24 index 2:
32/32 [==============================] - 0s 729us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.5934
16/16 [==============================] - 0s 1ms/step - loss: 1.1691
16/16 [==============================] - 0s 871us/step - loss: 1.2400
16/16 [==============================] - 0s 2ms/step - loss: 1.2445
16/16 [==============================] - 0s 2ms/step - loss: 1.2456
16/16 [==============================] - 0s 852us/step - loss: 1.2458
16/16 [==============================] - 0s 821us/step - loss: 1.2457
16/16 [==============================] - 0s 888us/step - loss: 1.2457
16/16 [==============================] - 0s 2ms/step - loss: 1.2456
16/16 [==============================] - 0s 848us/step - loss: 1.2456
Epoch 25 of 60

Testing for epoch 25 index 1:
32/32 [==============================] - 0s 614us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.5848
16/16 [==============================] - 0s 1ms/step - loss: 1.1934
16/16 [==============================] - 0s 2ms/step - loss: 1.2693
16/16 [==============================] - 0s 2ms/step - loss: 1.2739
16/16 [==============================] - 0s 1ms/step - loss: 1.2751
16/16 [==============================] - 0s 2ms/step - loss: 1.2753
16/16 [==============================] - 0s 2ms/step - loss: 1.2752
16/16 [==============================] - 0s 1ms/step - loss: 1.2751
16/16 [==============================] - 0s 874us/step - loss: 1.2750
16/16 [==============================] - 0s 822us/step - loss: 1.2750

Testing for epoch 25 index 2:
32/32 [==============================] - 0s 982us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.5871
16/16 [==============================] - 0s 2ms/step - loss: 1.2078
16/16 [==============================] - 0s 2ms/step - loss: 1.2829
16/16 [==============================] - 0s 2ms/step - loss: 1.2877
16/16 [==============================] - 0s 2ms/step - loss: 1.2888
16/16 [==============================] - 0s 2ms/step - loss: 1.2890
16/16 [==============================] - 0s 2ms/step - loss: 1.2890
16/16 [==============================] - 0s 2ms/step - loss: 1.2888
16/16 [==============================] - 0s 2ms/step - loss: 1.2888
16/16 [==============================] - 0s 888us/step - loss: 1.2888
Epoch 26 of 60

Testing for epoch 26 index 1:
32/32 [==============================] - 0s 606us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.5805
16/16 [==============================] - 0s 822us/step - loss: 1.2290
16/16 [==============================] - 0s 815us/step - loss: 1.3099
16/16 [==============================] - 0s 2ms/step - loss: 1.3149
16/16 [==============================] - 0s 790us/step - loss: 1.3161
16/16 [==============================] - 0s 815us/step - loss: 1.3163
16/16 [==============================] - 0s 2ms/step - loss: 1.3163
16/16 [==============================] - 0s 814us/step - loss: 1.3162
16/16 [==============================] - 0s 869us/step - loss: 1.3161
16/16 [==============================] - 0s 852us/step - loss: 1.3161

Testing for epoch 26 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.5851
16/16 [==============================] - 0s 1ms/step - loss: 1.2332
16/16 [==============================] - 0s 2ms/step - loss: 1.3115
16/16 [==============================] - 0s 824us/step - loss: 1.3164
16/16 [==============================] - 0s 1ms/step - loss: 1.3176
16/16 [==============================] - 0s 986us/step - loss: 1.3178
16/16 [==============================] - 0s 790us/step - loss: 1.3178
16/16 [==============================] - 0s 2ms/step - loss: 1.3177
16/16 [==============================] - 0s 2ms/step - loss: 1.3176
16/16 [==============================] - 0s 2ms/step - loss: 1.3176
Epoch 27 of 60

Testing for epoch 27 index 1:
32/32 [==============================] - 0s 574us/step
16/16 [==============================] - 0s 802us/step - loss: 0.5827
16/16 [==============================] - 0s 792us/step - loss: 1.2418
16/16 [==============================] - 0s 2ms/step - loss: 1.3207
16/16 [==============================] - 0s 774us/step - loss: 1.3257
16/16 [==============================] - 0s 807us/step - loss: 1.3269
16/16 [==============================] - 0s 791us/step - loss: 1.3271
16/16 [==============================] - 0s 964us/step - loss: 1.3270
16/16 [==============================] - 0s 2ms/step - loss: 1.3270
16/16 [==============================] - 0s 821us/step - loss: 1.3269
16/16 [==============================] - 0s 802us/step - loss: 1.3269

Testing for epoch 27 index 2:
32/32 [==============================] - 0s 580us/step
16/16 [==============================] - 0s 812us/step - loss: 0.5894
16/16 [==============================] - 0s 782us/step - loss: 1.2669
16/16 [==============================] - 0s 794us/step - loss: 1.3451
16/16 [==============================] - 0s 2ms/step - loss: 1.3501
16/16 [==============================] - 0s 843us/step - loss: 1.3513
16/16 [==============================] - 0s 792us/step - loss: 1.3515
16/16 [==============================] - 0s 773us/step - loss: 1.3515
16/16 [==============================] - 0s 802us/step - loss: 1.3514
16/16 [==============================] - 0s 2ms/step - loss: 1.3513
16/16 [==============================] - 0s 1ms/step - loss: 1.3513
Epoch 28 of 60

Testing for epoch 28 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 799us/step - loss: 0.5895
16/16 [==============================] - 0s 774us/step - loss: 1.2747
16/16 [==============================] - 0s 1ms/step - loss: 1.3543
16/16 [==============================] - 0s 861us/step - loss: 1.3594
16/16 [==============================] - 0s 2ms/step - loss: 1.3606
16/16 [==============================] - 0s 801us/step - loss: 1.3607
16/16 [==============================] - 0s 803us/step - loss: 1.3607
16/16 [==============================] - 0s 779us/step - loss: 1.3606
16/16 [==============================] - 0s 786us/step - loss: 1.3606
16/16 [==============================] - 0s 2ms/step - loss: 1.3605

Testing for epoch 28 index 2:
32/32 [==============================] - 0s 584us/step
16/16 [==============================] - 0s 820us/step - loss: 0.6016
16/16 [==============================] - 0s 2ms/step - loss: 1.2951
16/16 [==============================] - 0s 2ms/step - loss: 1.3737
16/16 [==============================] - 0s 2ms/step - loss: 1.3786
16/16 [==============================] - 0s 849us/step - loss: 1.3798
16/16 [==============================] - 0s 1ms/step - loss: 1.3800
16/16 [==============================] - 0s 853us/step - loss: 1.3799
16/16 [==============================] - 0s 980us/step - loss: 1.3798
16/16 [==============================] - 0s 2ms/step - loss: 1.3798
16/16 [==============================] - 0s 880us/step - loss: 1.3797
Epoch 29 of 60

Testing for epoch 29 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.6046
16/16 [==============================] - 0s 2ms/step - loss: 1.2864
16/16 [==============================] - 0s 1ms/step - loss: 1.3632
16/16 [==============================] - 0s 1ms/step - loss: 1.3680
16/16 [==============================] - 0s 804us/step - loss: 1.3690
16/16 [==============================] - 0s 2ms/step - loss: 1.3692
16/16 [==============================] - 0s 810us/step - loss: 1.3692
16/16 [==============================] - 0s 848us/step - loss: 1.3691
16/16 [==============================] - 0s 859us/step - loss: 1.3690
16/16 [==============================] - 0s 1ms/step - loss: 1.3690

Testing for epoch 29 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.6195
16/16 [==============================] - 0s 2ms/step - loss: 1.3102
16/16 [==============================] - 0s 2ms/step - loss: 1.3850
16/16 [==============================] - 0s 2ms/step - loss: 1.3898
16/16 [==============================] - 0s 2ms/step - loss: 1.3909
16/16 [==============================] - 0s 890us/step - loss: 1.3910
16/16 [==============================] - 0s 1ms/step - loss: 1.3910
16/16 [==============================] - 0s 789us/step - loss: 1.3909
16/16 [==============================] - 0s 783us/step - loss: 1.3908
16/16 [==============================] - 0s 773us/step - loss: 1.3908
Epoch 30 of 60

Testing for epoch 30 index 1:
32/32 [==============================] - 0s 608us/step
16/16 [==============================] - 0s 810us/step - loss: 0.6259
16/16 [==============================] - 0s 782us/step - loss: 1.3218
16/16 [==============================] - 0s 819us/step - loss: 1.3959
16/16 [==============================] - 0s 941us/step - loss: 1.4007
16/16 [==============================] - 0s 1ms/step - loss: 1.4018
16/16 [==============================] - 0s 2ms/step - loss: 1.4020
16/16 [==============================] - 0s 2ms/step - loss: 1.4020
16/16 [==============================] - 0s 2ms/step - loss: 1.4019
16/16 [==============================] - 0s 2ms/step - loss: 1.4018
16/16 [==============================] - 0s 2ms/step - loss: 1.4018

Testing for epoch 30 index 2:
32/32 [==============================] - 0s 568us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.6424
16/16 [==============================] - 0s 790us/step - loss: 1.3256
16/16 [==============================] - 0s 804us/step - loss: 1.3954
16/16 [==============================] - 0s 783us/step - loss: 1.4000
16/16 [==============================] - 0s 789us/step - loss: 1.4011
16/16 [==============================] - 0s 794us/step - loss: 1.4012
16/16 [==============================] - 0s 799us/step - loss: 1.4012
16/16 [==============================] - 0s 1ms/step - loss: 1.4011
16/16 [==============================] - 0s 2ms/step - loss: 1.4010
16/16 [==============================] - 0s 974us/step - loss: 1.4010
Epoch 31 of 60

Testing for epoch 31 index 1:
32/32 [==============================] - 0s 977us/step
16/16 [==============================] - 0s 1ms/step - loss: 0.6505
16/16 [==============================] - 0s 2ms/step - loss: 1.3458
16/16 [==============================] - 0s 2ms/step - loss: 1.4158
16/16 [==============================] - 0s 2ms/step - loss: 1.4204
16/16 [==============================] - 0s 2ms/step - loss: 1.4214
16/16 [==============================] - 0s 2ms/step - loss: 1.4216
16/16 [==============================] - 0s 2ms/step - loss: 1.4216
16/16 [==============================] - 0s 2ms/step - loss: 1.4215
16/16 [==============================] - 0s 2ms/step - loss: 1.4214
16/16 [==============================] - 0s 878us/step - loss: 1.4214

Testing for epoch 31 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.6688
16/16 [==============================] - 0s 2ms/step - loss: 1.3557
16/16 [==============================] - 0s 1ms/step - loss: 1.4234
16/16 [==============================] - 0s 1ms/step - loss: 1.4278
16/16 [==============================] - 0s 2ms/step - loss: 1.4288
16/16 [==============================] - 0s 2ms/step - loss: 1.4289
16/16 [==============================] - 0s 2ms/step - loss: 1.4289
16/16 [==============================] - 0s 825us/step - loss: 1.4288
16/16 [==============================] - 0s 2ms/step - loss: 1.4287
16/16 [==============================] - 0s 2ms/step - loss: 1.4287
Epoch 32 of 60

Testing for epoch 32 index 1:
32/32 [==============================] - 0s 564us/step
16/16 [==============================] - 0s 807us/step - loss: 0.6780
16/16 [==============================] - 0s 814us/step - loss: 1.3703
16/16 [==============================] - 0s 2ms/step - loss: 1.4374
16/16 [==============================] - 0s 2ms/step - loss: 1.4417
16/16 [==============================] - 0s 865us/step - loss: 1.4426
16/16 [==============================] - 0s 813us/step - loss: 1.4428
16/16 [==============================] - 0s 800us/step - loss: 1.4427
16/16 [==============================] - 0s 807us/step - loss: 1.4426
16/16 [==============================] - 0s 828us/step - loss: 1.4425
16/16 [==============================] - 0s 852us/step - loss: 1.4425

Testing for epoch 32 index 2:
32/32 [==============================] - 0s 591us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.6965
16/16 [==============================] - 0s 2ms/step - loss: 1.3779
16/16 [==============================] - 0s 2ms/step - loss: 1.4427
16/16 [==============================] - 0s 844us/step - loss: 1.4469
16/16 [==============================] - 0s 813us/step - loss: 1.4479
16/16 [==============================] - 0s 801us/step - loss: 1.4480
16/16 [==============================] - 0s 1ms/step - loss: 1.4480
16/16 [==============================] - 0s 2ms/step - loss: 1.4478
16/16 [==============================] - 0s 814us/step - loss: 1.4477
16/16 [==============================] - 0s 804us/step - loss: 1.4477
Epoch 33 of 60

Testing for epoch 33 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.7056
16/16 [==============================] - 0s 1ms/step - loss: 1.3908
16/16 [==============================] - 0s 989us/step - loss: 1.4549
16/16 [==============================] - 0s 2ms/step - loss: 1.4591
16/16 [==============================] - 0s 791us/step - loss: 1.4600
16/16 [==============================] - 0s 2ms/step - loss: 1.4602
16/16 [==============================] - 0s 1ms/step - loss: 1.4601
16/16 [==============================] - 0s 2ms/step - loss: 1.4600
16/16 [==============================] - 0s 2ms/step - loss: 1.4599
16/16 [==============================] - 0s 834us/step - loss: 1.4599

Testing for epoch 33 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.7234
16/16 [==============================] - 0s 1ms/step - loss: 1.3958
16/16 [==============================] - 0s 2ms/step - loss: 1.4578
16/16 [==============================] - 0s 2ms/step - loss: 1.4617
16/16 [==============================] - 0s 880us/step - loss: 1.4626
16/16 [==============================] - 0s 848us/step - loss: 1.4628
16/16 [==============================] - 0s 2ms/step - loss: 1.4627
16/16 [==============================] - 0s 782us/step - loss: 1.4626
16/16 [==============================] - 0s 758us/step - loss: 1.4625
16/16 [==============================] - 0s 814us/step - loss: 1.4625
Epoch 34 of 60

Testing for epoch 34 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.7344
16/16 [==============================] - 0s 841us/step - loss: 1.4243
16/16 [==============================] - 0s 805us/step - loss: 1.4870
16/16 [==============================] - 0s 2ms/step - loss: 1.4910
16/16 [==============================] - 0s 1ms/step - loss: 1.4919
16/16 [==============================] - 0s 1ms/step - loss: 1.4921
16/16 [==============================] - 0s 969us/step - loss: 1.4920
16/16 [==============================] - 0s 801us/step - loss: 1.4919
16/16 [==============================] - 0s 809us/step - loss: 1.4918
16/16 [==============================] - 0s 796us/step - loss: 1.4918

Testing for epoch 34 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 822us/step - loss: 0.7494
16/16 [==============================] - 0s 1ms/step - loss: 1.4233
16/16 [==============================] - 0s 846us/step - loss: 1.4838
16/16 [==============================] - 0s 1ms/step - loss: 1.4875
16/16 [==============================] - 0s 2ms/step - loss: 1.4883
16/16 [==============================] - 0s 2ms/step - loss: 1.4885
16/16 [==============================] - 0s 1ms/step - loss: 1.4884
16/16 [==============================] - 0s 2ms/step - loss: 1.4883
16/16 [==============================] - 0s 1ms/step - loss: 1.4882
16/16 [==============================] - 0s 2ms/step - loss: 1.4882
Epoch 35 of 60

Testing for epoch 35 index 1:
32/32 [==============================] - 0s 617us/step
16/16 [==============================] - 0s 780us/step - loss: 0.7594
16/16 [==============================] - 0s 785us/step - loss: 1.4475
16/16 [==============================] - 0s 786us/step - loss: 1.5084
16/16 [==============================] - 0s 790us/step - loss: 1.5122
16/16 [==============================] - 0s 2ms/step - loss: 1.5131
16/16 [==============================] - 0s 1ms/step - loss: 1.5132
16/16 [==============================] - 0s 2ms/step - loss: 1.5132
16/16 [==============================] - 0s 805us/step - loss: 1.5131
16/16 [==============================] - 0s 861us/step - loss: 1.5130
16/16 [==============================] - 0s 816us/step - loss: 1.5130

Testing for epoch 35 index 2:
32/32 [==============================] - 0s 574us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.7732
16/16 [==============================] - 0s 953us/step - loss: 1.4459
16/16 [==============================] - 0s 2ms/step - loss: 1.5047
16/16 [==============================] - 0s 1ms/step - loss: 1.5082
16/16 [==============================] - 0s 2ms/step - loss: 1.5090
16/16 [==============================] - 0s 2ms/step - loss: 1.5092
16/16 [==============================] - 0s 1ms/step - loss: 1.5091
16/16 [==============================] - 0s 2ms/step - loss: 1.5090
16/16 [==============================] - 0s 2ms/step - loss: 1.5089
16/16 [==============================] - 0s 2ms/step - loss: 1.5089
Epoch 36 of 60

Testing for epoch 36 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.7800
16/16 [==============================] - 0s 2ms/step - loss: 1.4588
16/16 [==============================] - 0s 2ms/step - loss: 1.5175
16/16 [==============================] - 0s 1ms/step - loss: 1.5209
16/16 [==============================] - 0s 832us/step - loss: 1.5218
16/16 [==============================] - 0s 2ms/step - loss: 1.5219
16/16 [==============================] - 0s 2ms/step - loss: 1.5219
16/16 [==============================] - 0s 869us/step - loss: 1.5217
16/16 [==============================] - 0s 875us/step - loss: 1.5216
16/16 [==============================] - 0s 791us/step - loss: 1.5216

Testing for epoch 36 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.7930
16/16 [==============================] - 0s 861us/step - loss: 1.4608
16/16 [==============================] - 0s 2ms/step - loss: 1.5175
16/16 [==============================] - 0s 932us/step - loss: 1.5209
16/16 [==============================] - 0s 827us/step - loss: 1.5217
16/16 [==============================] - 0s 2ms/step - loss: 1.5218
16/16 [==============================] - 0s 2ms/step - loss: 1.5218
16/16 [==============================] - 0s 2ms/step - loss: 1.5216
16/16 [==============================] - 0s 845us/step - loss: 1.5215
16/16 [==============================] - 0s 797us/step - loss: 1.5215
Epoch 37 of 60

Testing for epoch 37 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.8025
16/16 [==============================] - 0s 1ms/step - loss: 1.4900
16/16 [==============================] - 0s 815us/step - loss: 1.5481
16/16 [==============================] - 0s 2ms/step - loss: 1.5516
16/16 [==============================] - 0s 1ms/step - loss: 1.5524
16/16 [==============================] - 0s 802us/step - loss: 1.5526
16/16 [==============================] - 0s 2ms/step - loss: 1.5526
16/16 [==============================] - 0s 825us/step - loss: 1.5524
16/16 [==============================] - 0s 794us/step - loss: 1.5524
16/16 [==============================] - 0s 2ms/step - loss: 1.5523

Testing for epoch 37 index 2:
32/32 [==============================] - 0s 2ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.8189
16/16 [==============================] - 0s 1ms/step - loss: 1.5094
16/16 [==============================] - 0s 987us/step - loss: 1.5667
16/16 [==============================] - 0s 863us/step - loss: 1.5701
16/16 [==============================] - 0s 2ms/step - loss: 1.5710
16/16 [==============================] - 0s 2ms/step - loss: 1.5712
16/16 [==============================] - 0s 2ms/step - loss: 1.5711
16/16 [==============================] - 0s 2ms/step - loss: 1.5710
16/16 [==============================] - 0s 2ms/step - loss: 1.5709
16/16 [==============================] - 0s 938us/step - loss: 1.5709
Epoch 38 of 60

Testing for epoch 38 index 1:
32/32 [==============================] - 0s 745us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.8213
16/16 [==============================] - 0s 784us/step - loss: 1.5145
16/16 [==============================] - 0s 802us/step - loss: 1.5717
16/16 [==============================] - 0s 941us/step - loss: 1.5751
16/16 [==============================] - 0s 784us/step - loss: 1.5759
16/16 [==============================] - 0s 2ms/step - loss: 1.5761
16/16 [==============================] - 0s 823us/step - loss: 1.5760
16/16 [==============================] - 0s 956us/step - loss: 1.5759
16/16 [==============================] - 0s 2ms/step - loss: 1.5758
16/16 [==============================] - 0s 1ms/step - loss: 1.5758

Testing for epoch 38 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 840us/step - loss: 0.8398
16/16 [==============================] - 0s 813us/step - loss: 1.5413
16/16 [==============================] - 0s 2ms/step - loss: 1.5982
16/16 [==============================] - 0s 1ms/step - loss: 1.6017
16/16 [==============================] - 0s 2ms/step - loss: 1.6026
16/16 [==============================] - 0s 2ms/step - loss: 1.6027
16/16 [==============================] - 0s 2ms/step - loss: 1.6027
16/16 [==============================] - 0s 936us/step - loss: 1.6025
16/16 [==============================] - 0s 887us/step - loss: 1.6024
16/16 [==============================] - 0s 2ms/step - loss: 1.6024
Epoch 39 of 60

Testing for epoch 39 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 856us/step - loss: 0.8385
16/16 [==============================] - 0s 936us/step - loss: 1.5376
16/16 [==============================] - 0s 2ms/step - loss: 1.5940
16/16 [==============================] - 0s 2ms/step - loss: 1.5974
16/16 [==============================] - 0s 1ms/step - loss: 1.5982
16/16 [==============================] - 0s 2ms/step - loss: 1.5984
16/16 [==============================] - 0s 837us/step - loss: 1.5983
16/16 [==============================] - 0s 2ms/step - loss: 1.5982
16/16 [==============================] - 0s 1ms/step - loss: 1.5981
16/16 [==============================] - 0s 868us/step - loss: 1.5981

Testing for epoch 39 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.8519
16/16 [==============================] - 0s 923us/step - loss: 1.5491
16/16 [==============================] - 0s 2ms/step - loss: 1.6051
16/16 [==============================] - 0s 1ms/step - loss: 1.6085
16/16 [==============================] - 0s 2ms/step - loss: 1.6094
16/16 [==============================] - 0s 2ms/step - loss: 1.6095
16/16 [==============================] - 0s 2ms/step - loss: 1.6095
16/16 [==============================] - 0s 2ms/step - loss: 1.6093
16/16 [==============================] - 0s 879us/step - loss: 1.6093
16/16 [==============================] - 0s 832us/step - loss: 1.6092
Epoch 40 of 60

Testing for epoch 40 index 1:
32/32 [==============================] - 0s 2ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.8560
16/16 [==============================] - 0s 2ms/step - loss: 1.5641
16/16 [==============================] - 0s 849us/step - loss: 1.6209
16/16 [==============================] - 0s 886us/step - loss: 1.6243
16/16 [==============================] - 0s 2ms/step - loss: 1.6252
16/16 [==============================] - 0s 2ms/step - loss: 1.6253
16/16 [==============================] - 0s 872us/step - loss: 1.6253
16/16 [==============================] - 0s 1ms/step - loss: 1.6252
16/16 [==============================] - 0s 1ms/step - loss: 1.6251
16/16 [==============================] - 0s 2ms/step - loss: 1.6251

Testing for epoch 40 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 821us/step - loss: 0.8741
16/16 [==============================] - 0s 940us/step - loss: 1.5890
16/16 [==============================] - 0s 1ms/step - loss: 1.6459
16/16 [==============================] - 0s 1ms/step - loss: 1.6493
16/16 [==============================] - 0s 2ms/step - loss: 1.6501
16/16 [==============================] - 0s 914us/step - loss: 1.6503
16/16 [==============================] - 0s 810us/step - loss: 1.6502
16/16 [==============================] - 0s 811us/step - loss: 1.6501
16/16 [==============================] - 0s 2ms/step - loss: 1.6500
16/16 [==============================] - 0s 809us/step - loss: 1.6500
Epoch 41 of 60

Testing for epoch 41 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 799us/step - loss: 0.8705
16/16 [==============================] - 0s 2ms/step - loss: 1.5800
16/16 [==============================] - 0s 2ms/step - loss: 1.6359
16/16 [==============================] - 0s 824us/step - loss: 1.6392
16/16 [==============================] - 0s 2ms/step - loss: 1.6400
16/16 [==============================] - 0s 900us/step - loss: 1.6402
16/16 [==============================] - 0s 2ms/step - loss: 1.6401
16/16 [==============================] - 0s 800us/step - loss: 1.6400
16/16 [==============================] - 0s 759us/step - loss: 1.6399
16/16 [==============================] - 0s 774us/step - loss: 1.6399

Testing for epoch 41 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.8865
16/16 [==============================] - 0s 844us/step - loss: 1.5977
16/16 [==============================] - 0s 813us/step - loss: 1.6532
16/16 [==============================] - 0s 2ms/step - loss: 1.6564
16/16 [==============================] - 0s 811us/step - loss: 1.6572
16/16 [==============================] - 0s 782us/step - loss: 1.6573
16/16 [==============================] - 0s 772us/step - loss: 1.6573
16/16 [==============================] - 0s 1ms/step - loss: 1.6572
16/16 [==============================] - 0s 786us/step - loss: 1.6571
16/16 [==============================] - 0s 783us/step - loss: 1.6570
Epoch 42 of 60

Testing for epoch 42 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.8909
16/16 [==============================] - 0s 1ms/step - loss: 1.6130
16/16 [==============================] - 0s 1ms/step - loss: 1.6692
16/16 [==============================] - 0s 1ms/step - loss: 1.6724
16/16 [==============================] - 0s 811us/step - loss: 1.6733
16/16 [==============================] - 0s 2ms/step - loss: 1.6734
16/16 [==============================] - 0s 800us/step - loss: 1.6734
16/16 [==============================] - 0s 2ms/step - loss: 1.6733
16/16 [==============================] - 0s 805us/step - loss: 1.6732
16/16 [==============================] - 0s 760us/step - loss: 1.6732

Testing for epoch 42 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.9051
16/16 [==============================] - 0s 842us/step - loss: 1.6254
16/16 [==============================] - 0s 785us/step - loss: 1.6811
16/16 [==============================] - 0s 796us/step - loss: 1.6842
16/16 [==============================] - 0s 1ms/step - loss: 1.6851
16/16 [==============================] - 0s 2ms/step - loss: 1.6852
16/16 [==============================] - 0s 787us/step - loss: 1.6852
16/16 [==============================] - 0s 761us/step - loss: 1.6850
16/16 [==============================] - 0s 780us/step - loss: 1.6849
16/16 [==============================] - 0s 799us/step - loss: 1.6849
Epoch 43 of 60

Testing for epoch 43 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 828us/step - loss: 0.9117
16/16 [==============================] - 0s 951us/step - loss: 1.6455
16/16 [==============================] - 0s 2ms/step - loss: 1.7019
16/16 [==============================] - 0s 803us/step - loss: 1.7050
16/16 [==============================] - 0s 2ms/step - loss: 1.7059
16/16 [==============================] - 0s 2ms/step - loss: 1.7060
16/16 [==============================] - 0s 965us/step - loss: 1.7060
16/16 [==============================] - 0s 757us/step - loss: 1.7058
16/16 [==============================] - 0s 776us/step - loss: 1.7057
16/16 [==============================] - 0s 758us/step - loss: 1.7057

Testing for epoch 43 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 802us/step - loss: 0.9275
16/16 [==============================] - 0s 2ms/step - loss: 1.6609
16/16 [==============================] - 0s 843us/step - loss: 1.7168
16/16 [==============================] - 0s 785us/step - loss: 1.7199
16/16 [==============================] - 0s 800us/step - loss: 1.7208
16/16 [==============================] - 0s 1ms/step - loss: 1.7209
16/16 [==============================] - 0s 2ms/step - loss: 1.7209
16/16 [==============================] - 0s 855us/step - loss: 1.7208
16/16 [==============================] - 0s 812us/step - loss: 1.7207
16/16 [==============================] - 0s 803us/step - loss: 1.7207
Epoch 44 of 60

Testing for epoch 44 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 809us/step - loss: 0.9290
16/16 [==============================] - 0s 775us/step - loss: 1.6631
16/16 [==============================] - 0s 759us/step - loss: 1.7187
16/16 [==============================] - 0s 814us/step - loss: 1.7217
16/16 [==============================] - 0s 807us/step - loss: 1.7225
16/16 [==============================] - 0s 796us/step - loss: 1.7227
16/16 [==============================] - 0s 2ms/step - loss: 1.7226
16/16 [==============================] - 0s 1ms/step - loss: 1.7225
16/16 [==============================] - 0s 811us/step - loss: 1.7224
16/16 [==============================] - 0s 1ms/step - loss: 1.7224

Testing for epoch 44 index 2:
32/32 [==============================] - 0s 584us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.9455
16/16 [==============================] - 0s 1ms/step - loss: 1.6770
16/16 [==============================] - 0s 812us/step - loss: 1.7319
16/16 [==============================] - 0s 800us/step - loss: 1.7349
16/16 [==============================] - 0s 2ms/step - loss: 1.7358
16/16 [==============================] - 0s 2ms/step - loss: 1.7359
16/16 [==============================] - 0s 887us/step - loss: 1.7359
16/16 [==============================] - 0s 2ms/step - loss: 1.7358
16/16 [==============================] - 0s 2ms/step - loss: 1.7357
16/16 [==============================] - 0s 2ms/step - loss: 1.7357
Epoch 45 of 60

Testing for epoch 45 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 870us/step - loss: 0.9540
16/16 [==============================] - 0s 1ms/step - loss: 1.6976
16/16 [==============================] - 0s 2ms/step - loss: 1.7538
16/16 [==============================] - 0s 2ms/step - loss: 1.7569
16/16 [==============================] - 0s 2ms/step - loss: 1.7578
16/16 [==============================] - 0s 2ms/step - loss: 1.7580
16/16 [==============================] - 0s 2ms/step - loss: 1.7579
16/16 [==============================] - 0s 2ms/step - loss: 1.7578
16/16 [==============================] - 0s 2ms/step - loss: 1.7577
16/16 [==============================] - 0s 2ms/step - loss: 1.7577

Testing for epoch 45 index 2:
32/32 [==============================] - 0s 564us/step
16/16 [==============================] - 0s 787us/step - loss: 0.9649
16/16 [==============================] - 0s 790us/step - loss: 1.6943
16/16 [==============================] - 0s 2ms/step - loss: 1.7489
16/16 [==============================] - 0s 809us/step - loss: 1.7519
16/16 [==============================] - 0s 2ms/step - loss: 1.7527
16/16 [==============================] - 0s 2ms/step - loss: 1.7528
16/16 [==============================] - 0s 2ms/step - loss: 1.7528
16/16 [==============================] - 0s 2ms/step - loss: 1.7526
16/16 [==============================] - 0s 943us/step - loss: 1.7525
16/16 [==============================] - 0s 2ms/step - loss: 1.7525
Epoch 46 of 60

Testing for epoch 46 index 1:
32/32 [==============================] - 0s 876us/step
16/16 [==============================] - 0s 788us/step - loss: 0.9785
16/16 [==============================] - 0s 773us/step - loss: 1.7240
16/16 [==============================] - 0s 1ms/step - loss: 1.7800
16/16 [==============================] - 0s 2ms/step - loss: 1.7831
16/16 [==============================] - 0s 2ms/step - loss: 1.7839
16/16 [==============================] - 0s 859us/step - loss: 1.7840
16/16 [==============================] - 0s 773us/step - loss: 1.7840
16/16 [==============================] - 0s 2ms/step - loss: 1.7839
16/16 [==============================] - 0s 799us/step - loss: 1.7838
16/16 [==============================] - 0s 787us/step - loss: 1.7837

Testing for epoch 46 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.9940
16/16 [==============================] - 0s 2ms/step - loss: 1.7282
16/16 [==============================] - 0s 2ms/step - loss: 1.7828
16/16 [==============================] - 0s 982us/step - loss: 1.7858
16/16 [==============================] - 0s 894us/step - loss: 1.7866
16/16 [==============================] - 0s 809us/step - loss: 1.7867
16/16 [==============================] - 0s 2ms/step - loss: 1.7867
16/16 [==============================] - 0s 876us/step - loss: 1.7865
16/16 [==============================] - 0s 1ms/step - loss: 1.7864
16/16 [==============================] - 0s 2ms/step - loss: 1.7864
Epoch 47 of 60

Testing for epoch 47 index 1:
32/32 [==============================] - 0s 745us/step
16/16 [==============================] - 0s 2ms/step - loss: 1.0030
16/16 [==============================] - 0s 811us/step - loss: 1.7424
16/16 [==============================] - 0s 1ms/step - loss: 1.7977
16/16 [==============================] - 0s 2ms/step - loss: 1.8007
16/16 [==============================] - 0s 1ms/step - loss: 1.8015
16/16 [==============================] - 0s 827us/step - loss: 1.8017
16/16 [==============================] - 0s 2ms/step - loss: 1.8016
16/16 [==============================] - 0s 808us/step - loss: 1.8015
16/16 [==============================] - 0s 776us/step - loss: 1.8014
16/16 [==============================] - 0s 2ms/step - loss: 1.8014

Testing for epoch 47 index 2:
32/32 [==============================] - 0s 581us/step
16/16 [==============================] - 0s 784us/step - loss: 1.0184
16/16 [==============================] - 0s 2ms/step - loss: 1.7455
16/16 [==============================] - 0s 842us/step - loss: 1.7992
16/16 [==============================] - 0s 2ms/step - loss: 1.8021
16/16 [==============================] - 0s 827us/step - loss: 1.8028
16/16 [==============================] - 0s 2ms/step - loss: 1.8030
16/16 [==============================] - 0s 796us/step - loss: 1.8029
16/16 [==============================] - 0s 2ms/step - loss: 1.8028
16/16 [==============================] - 0s 2ms/step - loss: 1.8027
16/16 [==============================] - 0s 843us/step - loss: 1.8026
Epoch 48 of 60

Testing for epoch 48 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 791us/step - loss: 1.0233
16/16 [==============================] - 0s 788us/step - loss: 1.7481
16/16 [==============================] - 0s 772us/step - loss: 1.8019
16/16 [==============================] - 0s 2ms/step - loss: 1.8047
16/16 [==============================] - 0s 2ms/step - loss: 1.8055
16/16 [==============================] - 0s 2ms/step - loss: 1.8056
16/16 [==============================] - 0s 2ms/step - loss: 1.8056
16/16 [==============================] - 0s 2ms/step - loss: 1.8054
16/16 [==============================] - 0s 873us/step - loss: 1.8053
16/16 [==============================] - 0s 2ms/step - loss: 1.8053

Testing for epoch 48 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 843us/step - loss: 1.0415
16/16 [==============================] - 0s 879us/step - loss: 1.7551
16/16 [==============================] - 0s 1ms/step - loss: 1.8074
16/16 [==============================] - 0s 2ms/step - loss: 1.8102
16/16 [==============================] - 0s 1ms/step - loss: 1.8109
16/16 [==============================] - 0s 2ms/step - loss: 1.8110
16/16 [==============================] - 0s 2ms/step - loss: 1.8110
16/16 [==============================] - 0s 1ms/step - loss: 1.8108
16/16 [==============================] - 0s 2ms/step - loss: 1.8107
16/16 [==============================] - 0s 2ms/step - loss: 1.8107
Epoch 49 of 60

Testing for epoch 49 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 794us/step - loss: 1.0581
16/16 [==============================] - 0s 1ms/step - loss: 1.7845
16/16 [==============================] - 0s 864us/step - loss: 1.8379
16/16 [==============================] - 0s 913us/step - loss: 1.8407
16/16 [==============================] - 0s 2ms/step - loss: 1.8415
16/16 [==============================] - 0s 830us/step - loss: 1.8416
16/16 [==============================] - 0s 2ms/step - loss: 1.8416
16/16 [==============================] - 0s 950us/step - loss: 1.8414
16/16 [==============================] - 0s 1ms/step - loss: 1.8413
16/16 [==============================] - 0s 2ms/step - loss: 1.8413

Testing for epoch 49 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 972us/step - loss: 1.0727
16/16 [==============================] - 0s 2ms/step - loss: 1.7840
16/16 [==============================] - 0s 955us/step - loss: 1.8360
16/16 [==============================] - 0s 792us/step - loss: 1.8388
16/16 [==============================] - 0s 2ms/step - loss: 1.8395
16/16 [==============================] - 0s 2ms/step - loss: 1.8397
16/16 [==============================] - 0s 2ms/step - loss: 1.8396
16/16 [==============================] - 0s 894us/step - loss: 1.8395
16/16 [==============================] - 0s 2ms/step - loss: 1.8394
16/16 [==============================] - 0s 2ms/step - loss: 1.8394
Epoch 50 of 60

Testing for epoch 50 index 1:
32/32 [==============================] - 0s 612us/step
16/16 [==============================] - 0s 814us/step - loss: 1.0911
16/16 [==============================] - 0s 799us/step - loss: 1.8158
16/16 [==============================] - 0s 792us/step - loss: 1.8689
16/16 [==============================] - 0s 832us/step - loss: 1.8717
16/16 [==============================] - 0s 2ms/step - loss: 1.8724
16/16 [==============================] - 0s 2ms/step - loss: 1.8725
16/16 [==============================] - 0s 805us/step - loss: 1.8725
16/16 [==============================] - 0s 2ms/step - loss: 1.8724
16/16 [==============================] - 0s 2ms/step - loss: 1.8723
16/16 [==============================] - 0s 2ms/step - loss: 1.8722

Testing for epoch 50 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 852us/step - loss: 1.1092
16/16 [==============================] - 0s 2ms/step - loss: 1.8231
16/16 [==============================] - 0s 788us/step - loss: 1.8750
16/16 [==============================] - 0s 805us/step - loss: 1.8778
16/16 [==============================] - 0s 2ms/step - loss: 1.8785
16/16 [==============================] - 0s 884us/step - loss: 1.8786
16/16 [==============================] - 0s 2ms/step - loss: 1.8786
16/16 [==============================] - 0s 2ms/step - loss: 1.8785
16/16 [==============================] - 0s 853us/step - loss: 1.8784
16/16 [==============================] - 0s 805us/step - loss: 1.8784
Epoch 51 of 60

Testing for epoch 51 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 848us/step - loss: 1.1099
16/16 [==============================] - 0s 2ms/step - loss: 1.8169
16/16 [==============================] - 0s 2ms/step - loss: 1.8684
16/16 [==============================] - 0s 2ms/step - loss: 1.8711
16/16 [==============================] - 0s 894us/step - loss: 1.8718
16/16 [==============================] - 0s 1ms/step - loss: 1.8720
16/16 [==============================] - 0s 1ms/step - loss: 1.8719
16/16 [==============================] - 0s 1ms/step - loss: 1.8718
16/16 [==============================] - 0s 1ms/step - loss: 1.8718
16/16 [==============================] - 0s 2ms/step - loss: 1.8717

Testing for epoch 51 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 1.1318
16/16 [==============================] - 0s 939us/step - loss: 1.8340
16/16 [==============================] - 0s 2ms/step - loss: 1.8847
16/16 [==============================] - 0s 904us/step - loss: 1.8874
16/16 [==============================] - 0s 854us/step - loss: 1.8880
16/16 [==============================] - 0s 894us/step - loss: 1.8882
16/16 [==============================] - 0s 2ms/step - loss: 1.8882
16/16 [==============================] - 0s 981us/step - loss: 1.8880
16/16 [==============================] - 0s 2ms/step - loss: 1.8879
16/16 [==============================] - 0s 951us/step - loss: 1.8879
Epoch 52 of 60

Testing for epoch 52 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 933us/step - loss: 1.1444
16/16 [==============================] - 0s 1ms/step - loss: 1.8546
16/16 [==============================] - 0s 2ms/step - loss: 1.9062
16/16 [==============================] - 0s 2ms/step - loss: 1.9089
16/16 [==============================] - 0s 2ms/step - loss: 1.9096
16/16 [==============================] - 0s 2ms/step - loss: 1.9097
16/16 [==============================] - 0s 2ms/step - loss: 1.9097
16/16 [==============================] - 0s 935us/step - loss: 1.9096
16/16 [==============================] - 0s 2ms/step - loss: 1.9095
16/16 [==============================] - 0s 2ms/step - loss: 1.9095

Testing for epoch 52 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 1.1574
16/16 [==============================] - 0s 912us/step - loss: 1.8545
16/16 [==============================] - 0s 1ms/step - loss: 1.9046
16/16 [==============================] - 0s 2ms/step - loss: 1.9072
16/16 [==============================] - 0s 779us/step - loss: 1.9078
16/16 [==============================] - 0s 782us/step - loss: 1.9079
16/16 [==============================] - 0s 1ms/step - loss: 1.9079
16/16 [==============================] - 0s 994us/step - loss: 1.9077
16/16 [==============================] - 0s 2ms/step - loss: 1.9076
16/16 [==============================] - 0s 2ms/step - loss: 1.9076
Epoch 53 of 60

Testing for epoch 53 index 1:
32/32 [==============================] - 0s 920us/step
16/16 [==============================] - 0s 2ms/step - loss: 1.1652
16/16 [==============================] - 0s 2ms/step - loss: 1.8650
16/16 [==============================] - 0s 2ms/step - loss: 1.9154
16/16 [==============================] - 0s 2ms/step - loss: 1.9180
16/16 [==============================] - 0s 919us/step - loss: 1.9186
16/16 [==============================] - 0s 1ms/step - loss: 1.9187
16/16 [==============================] - 0s 935us/step - loss: 1.9187
16/16 [==============================] - 0s 2ms/step - loss: 1.9185
16/16 [==============================] - 0s 897us/step - loss: 1.9184
16/16 [==============================] - 0s 836us/step - loss: 1.9184

Testing for epoch 53 index 2:
32/32 [==============================] - 0s 647us/step
16/16 [==============================] - 0s 812us/step - loss: 1.1824
16/16 [==============================] - 0s 769us/step - loss: 1.8743
16/16 [==============================] - 0s 2ms/step - loss: 1.9238
16/16 [==============================] - 0s 796us/step - loss: 1.9264
16/16 [==============================] - 0s 2ms/step - loss: 1.9270
16/16 [==============================] - 0s 812us/step - loss: 1.9272
16/16 [==============================] - 0s 813us/step - loss: 1.9271
16/16 [==============================] - 0s 1ms/step - loss: 1.9270
16/16 [==============================] - 0s 856us/step - loss: 1.9269
16/16 [==============================] - 0s 810us/step - loss: 1.9268
Epoch 54 of 60

Testing for epoch 54 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 844us/step - loss: 1.1988
16/16 [==============================] - 0s 1ms/step - loss: 1.9035
16/16 [==============================] - 0s 932us/step - loss: 1.9535
16/16 [==============================] - 0s 787us/step - loss: 1.9561
16/16 [==============================] - 0s 792us/step - loss: 1.9568
16/16 [==============================] - 0s 794us/step - loss: 1.9570
16/16 [==============================] - 0s 1ms/step - loss: 1.9569
16/16 [==============================] - 0s 814us/step - loss: 1.9568
16/16 [==============================] - 0s 1ms/step - loss: 1.9567
16/16 [==============================] - 0s 2ms/step - loss: 1.9567

Testing for epoch 54 index 2:
32/32 [==============================] - 0s 553us/step
16/16 [==============================] - 0s 2ms/step - loss: 1.2127
16/16 [==============================] - 0s 822us/step - loss: 1.9062
16/16 [==============================] - 0s 811us/step - loss: 1.9551
16/16 [==============================] - 0s 2ms/step - loss: 1.9577
16/16 [==============================] - 0s 803us/step - loss: 1.9584
16/16 [==============================] - 0s 798us/step - loss: 1.9585
16/16 [==============================] - 0s 773us/step - loss: 1.9585
16/16 [==============================] - 0s 796us/step - loss: 1.9584
16/16 [==============================] - 0s 806us/step - loss: 1.9583
16/16 [==============================] - 0s 808us/step - loss: 1.9583
Epoch 55 of 60

Testing for epoch 55 index 1:
32/32 [==============================] - 0s 636us/step
16/16 [==============================] - 0s 897us/step - loss: 1.2206
16/16 [==============================] - 0s 805us/step - loss: 1.9180
16/16 [==============================] - 0s 787us/step - loss: 1.9664
16/16 [==============================] - 0s 1ms/step - loss: 1.9689
16/16 [==============================] - 0s 2ms/step - loss: 1.9695
16/16 [==============================] - 0s 800us/step - loss: 1.9697
16/16 [==============================] - 0s 1ms/step - loss: 1.9696
16/16 [==============================] - 0s 2ms/step - loss: 1.9695
16/16 [==============================] - 0s 913us/step - loss: 1.9694
16/16 [==============================] - 0s 2ms/step - loss: 1.9694

Testing for epoch 55 index 2:
32/32 [==============================] - 0s 563us/step
16/16 [==============================] - 0s 809us/step - loss: 1.2428
16/16 [==============================] - 0s 839us/step - loss: 1.9369
16/16 [==============================] - 0s 818us/step - loss: 1.9851
16/16 [==============================] - 0s 1ms/step - loss: 1.9876
16/16 [==============================] - 0s 2ms/step - loss: 1.9883
16/16 [==============================] - 0s 2ms/step - loss: 1.9884
16/16 [==============================] - 0s 1ms/step - loss: 1.9884
16/16 [==============================] - 0s 1ms/step - loss: 1.9883
16/16 [==============================] - 0s 2ms/step - loss: 1.9882
16/16 [==============================] - 0s 1ms/step - loss: 1.9882
Epoch 56 of 60

Testing for epoch 56 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 1.2530
16/16 [==============================] - 0s 2ms/step - loss: 1.9526
16/16 [==============================] - 0s 1ms/step - loss: 2.0011
16/16 [==============================] - 0s 892us/step - loss: 2.0036
16/16 [==============================] - 0s 805us/step - loss: 2.0042
16/16 [==============================] - 0s 782us/step - loss: 2.0044
16/16 [==============================] - 0s 788us/step - loss: 2.0043
16/16 [==============================] - 0s 2ms/step - loss: 2.0042
16/16 [==============================] - 0s 2ms/step - loss: 2.0041
16/16 [==============================] - 0s 854us/step - loss: 2.0040

Testing for epoch 56 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 1.2613
16/16 [==============================] - 0s 2ms/step - loss: 1.9478
16/16 [==============================] - 0s 2ms/step - loss: 1.9951
16/16 [==============================] - 0s 2ms/step - loss: 1.9976
16/16 [==============================] - 0s 882us/step - loss: 1.9982
16/16 [==============================] - 0s 1ms/step - loss: 1.9984
16/16 [==============================] - 0s 1ms/step - loss: 1.9984
16/16 [==============================] - 0s 793us/step - loss: 1.9982
16/16 [==============================] - 0s 2ms/step - loss: 1.9981
16/16 [==============================] - 0s 821us/step - loss: 1.9981
Epoch 57 of 60

Testing for epoch 57 index 1:
32/32 [==============================] - 0s 621us/step
16/16 [==============================] - 0s 1ms/step - loss: 1.2723
16/16 [==============================] - 0s 2ms/step - loss: 1.9662
16/16 [==============================] - 0s 2ms/step - loss: 2.0142
16/16 [==============================] - 0s 1ms/step - loss: 2.0167
16/16 [==============================] - 0s 821us/step - loss: 2.0173
16/16 [==============================] - 0s 789us/step - loss: 2.0175
16/16 [==============================] - 0s 2ms/step - loss: 2.0174
16/16 [==============================] - 0s 803us/step - loss: 2.0173
16/16 [==============================] - 0s 2ms/step - loss: 2.0172
16/16 [==============================] - 0s 2ms/step - loss: 2.0172

Testing for epoch 57 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 835us/step - loss: 1.2890
16/16 [==============================] - 0s 2ms/step - loss: 1.9763
16/16 [==============================] - 0s 2ms/step - loss: 2.0236
16/16 [==============================] - 0s 2ms/step - loss: 2.0261
16/16 [==============================] - 0s 1ms/step - loss: 2.0267
16/16 [==============================] - 0s 910us/step - loss: 2.0268
16/16 [==============================] - 0s 885us/step - loss: 2.0268
16/16 [==============================] - 0s 2ms/step - loss: 2.0267
16/16 [==============================] - 0s 974us/step - loss: 2.0266
16/16 [==============================] - 0s 864us/step - loss: 2.0266
Epoch 58 of 60

Testing for epoch 58 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 802us/step - loss: 1.2869
16/16 [==============================] - 0s 1ms/step - loss: 1.9705
16/16 [==============================] - 0s 2ms/step - loss: 2.0173
16/16 [==============================] - 0s 2ms/step - loss: 2.0197
16/16 [==============================] - 0s 2ms/step - loss: 2.0203
16/16 [==============================] - 0s 1ms/step - loss: 2.0204
16/16 [==============================] - 0s 2ms/step - loss: 2.0203
16/16 [==============================] - 0s 2ms/step - loss: 2.0202
16/16 [==============================] - 0s 2ms/step - loss: 2.0201
16/16 [==============================] - 0s 833us/step - loss: 2.0201

Testing for epoch 58 index 2:
32/32 [==============================] - 0s 584us/step
16/16 [==============================] - 0s 791us/step - loss: 1.3158
16/16 [==============================] - 0s 818us/step - loss: 2.0032
16/16 [==============================] - 0s 785us/step - loss: 2.0502
16/16 [==============================] - 0s 1ms/step - loss: 2.0527
16/16 [==============================] - 0s 1ms/step - loss: 2.0533
16/16 [==============================] - 0s 847us/step - loss: 2.0535
16/16 [==============================] - 0s 790us/step - loss: 2.0535
16/16 [==============================] - 0s 2ms/step - loss: 2.0534
16/16 [==============================] - 0s 831us/step - loss: 2.0533
16/16 [==============================] - 0s 2ms/step - loss: 2.0533
Epoch 59 of 60

Testing for epoch 59 index 1:
32/32 [==============================] - 0s 591us/step
16/16 [==============================] - 0s 803us/step - loss: 1.3201
16/16 [==============================] - 0s 811us/step - loss: 2.0086
16/16 [==============================] - 0s 783us/step - loss: 2.0555
16/16 [==============================] - 0s 792us/step - loss: 2.0580
16/16 [==============================] - 0s 2ms/step - loss: 2.0585
16/16 [==============================] - 0s 828us/step - loss: 2.0587
16/16 [==============================] - 0s 2ms/step - loss: 2.0586
16/16 [==============================] - 0s 2ms/step - loss: 2.0585
16/16 [==============================] - 0s 803us/step - loss: 2.0584
16/16 [==============================] - 0s 778us/step - loss: 2.0584

Testing for epoch 59 index 2:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 853us/step - loss: 1.3376
16/16 [==============================] - 0s 2ms/step - loss: 2.0187
16/16 [==============================] - 0s 891us/step - loss: 2.0651
16/16 [==============================] - 0s 2ms/step - loss: 2.0675
16/16 [==============================] - 0s 2ms/step - loss: 2.0681
16/16 [==============================] - 0s 2ms/step - loss: 2.0683
16/16 [==============================] - 0s 1ms/step - loss: 2.0682
16/16 [==============================] - 0s 2ms/step - loss: 2.0681
16/16 [==============================] - 0s 2ms/step - loss: 2.0680
16/16 [==============================] - 0s 2ms/step - loss: 2.0680
Epoch 60 of 60

Testing for epoch 60 index 1:
32/32 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 1.3388
16/16 [==============================] - 0s 830us/step - loss: 2.0171
16/16 [==============================] - 0s 1ms/step - loss: 2.0632
16/16 [==============================] - 0s 788us/step - loss: 2.0656
16/16 [==============================] - 0s 847us/step - loss: 2.0662
16/16 [==============================] - 0s 1ms/step - loss: 2.0663
16/16 [==============================] - 0s 2ms/step - loss: 2.0662
16/16 [==============================] - 0s 948us/step - loss: 2.0661
16/16 [==============================] - 0s 2ms/step - loss: 2.0660
16/16 [==============================] - 0s 2ms/step - loss: 2.0660

Testing for epoch 60 index 2:
32/32 [==============================] - 0s 585us/step
16/16 [==============================] - 0s 791us/step - loss: 1.3679
16/16 [==============================] - 0s 2ms/step - loss: 2.0481
16/16 [==============================] - 0s 791us/step - loss: 2.0943
16/16 [==============================] - 0s 785us/step - loss: 2.0968
16/16 [==============================] - 0s 784us/step - loss: 2.0974
16/16 [==============================] - 0s 788us/step - loss: 2.0976
16/16 [==============================] - 0s 823us/step - loss: 2.0975
16/16 [==============================] - 0s 786us/step - loss: 2.0974
16/16 [==============================] - 0s 803us/step - loss: 2.0973
16/16 [==============================] - 0s 812us/step - loss: 2.0973
32/32 [==============================] - 0s 572us/step

MO_GAAL(contamination=0.05, k=10, lr_d=0.01, lr_g=0.0001, momentum=0.9,
    stop_epochs=20)

outlier_MO_GAAL_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_orbit,outlier_MO_GAAL_one)
_conf.conf("MO-GAAL (Liu et al., 2019)")

Accuracy: 0.951
Precision: 0.000
Recall: 0.000
F1 Score: 0.000

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/metrics/_classification.py:1469: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.
  _warn_prf(average, modifier, msg_start, len(result))

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.951
Precision(TP / TP + FP): nan
Recall(TP / TP + FN): 0.000
F1 Score(2*precision*recall/precision+recall): nan

/tmp/ipykernel_3852735/4166638268.py:3: RuntimeWarning: invalid value encountered in long_scalars
  print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
/tmp/ipykernel_3852735/4166638268.py:5: RuntimeWarning: invalid value encountered in long_scalars
  print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

fpr, tpr, thresh = roc_curve(outlier_true_orbit,clf.decision_function(_df))

32/32 [==============================] - 0s 557us/step

auc(fpr, tpr)

0.5095924805253332

tab_orbit = pd.concat([tab_orbit,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_orbit

	Accuracy	Precision	Recall	F1	AUC
GODE	0.961	0.600000	0.612245	0.606061	0.893023
LOF (Breunig et al., 2000)	0.921	0.200000	0.204082	0.202020	0.664124
kNN (Ramaswamy et al., 2000)	0.947	0.460000	0.469388	0.464646	0.847990
CBLOF (He et al., 2003)	0.911	0.100000	0.102041	0.101010	0.533402
OCSVM (Sch ̈olkopf et al., 2001)	0.893	0.210000	0.428571	0.281879	0.788751
MCD (Hardin and Rocke, 2004)	0.911	0.100000	0.102041	0.101010	0.454001
Feature Bagging (Lazarevic and Kumar, 2005)	0.921	0.200000	0.204082	0.202020	0.677997
ABOD (Kriegel et al., 2008)	0.951	0.500000	0.510204	0.505051	0.863924
Isolation Forest (Liu et al., 2008)	0.925	0.240000	0.244898	0.242424	0.618018
HBOS (Goldstein and Dengel, 2012)	0.921	0.105263	0.081633	0.091954	0.529743
SOS (Janssens et al., 2012)	0.941	0.400000	0.408163	0.404040	0.844224
SO-GAAL (Liu et al., 2019)	0.951	0.000000	0.000000	0.000000	0.464345
MO-GAAL (Liu et al., 2019)	0.951	0.000000	0.000000	0.000000	0.509592

LSCP_Orbit

detectors = [KNN(), LOF(), OCSVM()]
clf = LSCP(detectors,contamination=0.05, random_state=77)
clf.fit(_df[['x', 'y','f']])

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/pyod/models/lscp.py:382: UserWarning: The number of histogram bins is greater than the number of classifiers, reducing n_bins to n_clf.
  warnings.warn(

LSCP(contamination=0.05,
   detector_list=[KNN(algorithm='auto', contamination=0.1, leaf_size=30, method='largest',
  metric='minkowski', metric_params=None, n_jobs=1, n_neighbors=5, p=2,
  radius=1.0), LOF(algorithm='auto', contamination=0.1, leaf_size=30, metric='minkowski',
  metric_params=None, n_jobs=1, n_neighbors=20, novelty=True, p=2), OCSVM(cache_size=200, coef0=0.0, contamination=0.1, degree=3, gamma='auto',
   kernel='rbf', max_iter=-1, nu=0.5, shrinking=True, tol=0.001,
   verbose=False)],
   local_max_features=1.0, local_region_size=30, n_bins=3,
   random_state=RandomState(MT19937) at 0x7FD5F8280640)

outlier_LSCP_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_orbit,outlier_LSCP_one)
_conf.conf("LSCP (Zhao et al., 2019)")

Accuracy: 0.947
Precision: 0.460
Recall: 0.469
F1 Score: 0.465

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 0.947
Precision(TP / TP + FP): 0.460
Recall(TP / TP + FN): 0.469
F1 Score(2*precision*recall/precision+recall): 0.465

fpr, tpr, thresh = roc_curve(outlier_true_orbit,clf.decision_function(_df))

auc(fpr, tpr)

0.849288611343591

tab_orbit = pd.concat([tab_orbit,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_orbit

	Accuracy	Precision	Recall	F1	AUC
GODE	0.961	0.600000	0.612245	0.606061	0.893023
LOF (Breunig et al., 2000)	0.921	0.200000	0.204082	0.202020	0.664124
kNN (Ramaswamy et al., 2000)	0.947	0.460000	0.469388	0.464646	0.847990
CBLOF (He et al., 2003)	0.911	0.100000	0.102041	0.101010	0.533402
OCSVM (Sch ̈olkopf et al., 2001)	0.893	0.210000	0.428571	0.281879	0.788751
MCD (Hardin and Rocke, 2004)	0.911	0.100000	0.102041	0.101010	0.454001
Feature Bagging (Lazarevic and Kumar, 2005)	0.921	0.200000	0.204082	0.202020	0.677997
ABOD (Kriegel et al., 2008)	0.951	0.500000	0.510204	0.505051	0.863924
Isolation Forest (Liu et al., 2008)	0.925	0.240000	0.244898	0.242424	0.618018
HBOS (Goldstein and Dengel, 2012)	0.921	0.105263	0.081633	0.091954	0.529743
SOS (Janssens et al., 2012)	0.941	0.400000	0.408163	0.404040	0.844224
SO-GAAL (Liu et al., 2019)	0.951	0.000000	0.000000	0.000000	0.464345
MO-GAAL (Liu et al., 2019)	0.951	0.000000	0.000000	0.000000	0.509592
LSCP (Zhao et al., 2019)	0.947	0.460000	0.469388	0.464646	0.849289

tab_orbit

round(tab_orbit,3)

	Accuracy	Precision	Recall	F1	AUC
GODE	0.961	0.600	0.612	0.606	0.893
LOF (Breunig et al., 2000)	0.921	0.200	0.204	0.202	0.664
kNN (Ramaswamy et al., 2000)	0.947	0.460	0.469	0.465	0.848
CBLOF (He et al., 2003)	0.911	0.100	0.102	0.101	0.533
OCSVM (Sch ̈olkopf et al., 2001)	0.893	0.210	0.429	0.282	0.789
MCD (Hardin and Rocke, 2004)	0.911	0.100	0.102	0.101	0.454
Feature Bagging (Lazarevic and Kumar, 2005)	0.921	0.200	0.204	0.202	0.678
ABOD (Kriegel et al., 2008)	0.951	0.500	0.510	0.505	0.864
Isolation Forest (Liu et al., 2008)	0.925	0.240	0.245	0.242	0.618
HBOS (Goldstein and Dengel, 2012)	0.921	0.105	0.082	0.092	0.530
SOS (Janssens et al., 2012)	0.941	0.400	0.408	0.404	0.844
SO-GAAL (Liu et al., 2019)	0.951	0.000	0.000	0.000	0.464
MO-GAAL (Liu et al., 2019)	0.951	0.000	0.000	0.000	0.510
LSCP (Zhao et al., 2019)	0.947	0.460	0.469	0.465	0.849

Bunny

G = graphs.Bunny()
n = G.N
g = filters.Heat(G, tau=75) 
n=2503
np.random.seed(1212)
normal = np.around(np.random.normal(size=n),15)
unif = np.concatenate([np.random.uniform(low=3,high=7,size=60), np.random.uniform(low=-7,high=-3,size=60),np.zeros(n-120)]); np.random.shuffle(unif)
noise = normal + unif
f = np.zeros(n)
f[1000] = -3234
f = g.filter(f, method='chebyshev') 

2023-11-27 13:28:42,553:[WARNING](pygsp.graphs.graph.lmax): The largest eigenvalue G.lmax is not available, we need to estimate it. Explicitly call G.estimate_lmax() or G.compute_fourier_basis() once beforehand to suppress the warning.

G.coords.shape
_W = G.W.toarray()
_x = G.coords[:,0]
_y = G.coords[:,1]
_z = -G.coords[:,2]
_df1 = {'W':_W,'x':_x,'y':_y,'z':_z, 'fnoise':f+noise,'f' : f, 'noise': noise,'unif':unif}
_df = pd.DataFrame({'x': _df1['x'],'y':_df1['y'],'z':_df1['z'],'fnoise':_df1['fnoise'],'f':_df1['f'],'noise':_df1['noise']})
unif = _df1['unif']
outlier_true_bunny = unif.copy()
outlier_true_bunny = list(map(lambda x: 1 if x !=0  else 0,outlier_true_bunny))
X = np.array(_df)[:,:4]

GODE_Bunny

_W = _df1['W']
_BUNNY = BUNNY(_df)
_BUNNY.fit(sd=20,ref=10)

outlier_GODE_one_old = (_BUNNY.df['Residual']**2).tolist()
sorted_data = sorted(outlier_GODE_one_old,reverse=True)
index = int(len(sorted_data) * 0.05)
five_percent = sorted_data[index]
outlier_GODE_one = list(map(lambda x: 1 if x > five_percent else 0,outlier_GODE_one_old))

_conf = Conf_matrx(outlier_true_bunny,outlier_GODE_one)
_conf.conf("GODE")

Accuracy: 0.988
Precision: 0.864
Recall: 0.900
F1 Score: 0.882

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 2.474
Precision(TP / TP + FP): 0.864
Recall(TP / TP + FN): 0.900
F1 Score(2*precision*recall/precision+recall): 0.882

fpr, tpr, thresh = roc_curve(outlier_true_bunny,outlier_GODE_one_old)

auc(fpr, tpr)

0.9962267449993005

tab_bunny = pd.concat([tab_bunny,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_bunny

	Accuracy	Precision	Recall	F1	AUC
GODE	0.988414	0.864	0.9	0.881633	0.996227

LOF_Bunny

np.random.seed(77)
clf = LOF(contamination=0.05)
clf.fit(_df[['x', 'y','fnoise']])

LOF(algorithm='auto', contamination=0.05, leaf_size=30, metric='minkowski',
  metric_params=None, n_jobs=1, n_neighbors=20, novelty=True, p=2)

outlier_LOF_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_bunny,clf.fit_predict(X))
_conf.conf("LOF (Breunig et al., 2000)")

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/utils/deprecation.py:86: FutureWarning: Function fit_predict is deprecated
  warnings.warn(msg, category=FutureWarning)

Accuracy: 0.943
Precision: 0.413
Recall: 0.433
F1 Score: 0.423

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 2.361
Precision(TP / TP + FP): 0.413
Recall(TP / TP + FN): 0.433
F1 Score(2*precision*recall/precision+recall): 0.423

fpr, tpr, thresh = roc_curve(outlier_true_bunny,clf.decision_function(X))

auc(fpr, tpr)

0.8191110644845433

tab_bunny = pd.concat([tab_bunny,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_bunny

	Accuracy	Precision	Recall	F1	AUC
GODE	0.988414	0.864000	0.900000	0.881633	0.996227
LOF (Breunig et al., 2000)	0.943268	0.412698	0.433333	0.422764	0.819111

KNN_Bunny

np.random.seed(77)
clf = KNN(contamination=0.05)
clf.fit(_df[['x', 'y','fnoise']])

KNN(algorithm='auto', contamination=0.05, leaf_size=30, method='largest',
  metric='minkowski', metric_params=None, n_jobs=1, n_neighbors=5, p=2,
  radius=1.0)

outlier_KNN_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_bunny,outlier_KNN_one)
_conf.conf("kNN (Ramaswamy et al., 2000)")

Accuracy: 0.987
Precision: 0.849
Recall: 0.892
F1 Score: 0.870

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 2.471
Precision(TP / TP + FP): 0.849
Recall(TP / TP + FN): 0.892
F1 Score(2*precision*recall/precision+recall): 0.870

fpr, tpr, thresh = roc_curve(outlier_true_bunny,clf.decision_function(_df[['x', 'y','fnoise']]))

auc(fpr, tpr)

0.984438382990628

tab_bunny = pd.concat([tab_bunny,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_bunny

	Accuracy	Precision	Recall	F1	AUC
GODE	0.988414	0.864000	0.900000	0.881633	0.996227
LOF (Breunig et al., 2000)	0.943268	0.412698	0.433333	0.422764	0.819111
kNN (Ramaswamy et al., 2000)	0.987215	0.849206	0.891667	0.869919	0.984438

CBLOF_Bunny

clf = CBLOF(contamination=0.05,random_state=77)
clf.fit(_df[['x', 'y','fnoise']])

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning
  super()._check_params_vs_input(X, default_n_init=10)

CBLOF(alpha=0.9, beta=5, check_estimator=False, clustering_estimator=None,
   contamination=0.05, n_clusters=8, n_jobs=None, random_state=77,
   use_weights=False)

outlier_CBLOF_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_bunny,outlier_CBLOF_one)
_conf.conf("CBLOF (He et al., 2003)")

Accuracy: 0.981
Precision: 0.786
Recall: 0.825
F1 Score: 0.805

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 2.455
Precision(TP / TP + FP): 0.786
Recall(TP / TP + FN): 0.825
F1 Score(2*precision*recall/precision+recall): 0.805

fpr, tpr, thresh = roc_curve(outlier_true_bunny,clf.decision_function(_df[['x', 'y','fnoise']]))

auc(fpr, tpr)

0.9705448314449574

tab_bunny = pd.concat([tab_bunny,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_bunny

	Accuracy	Precision	Recall	F1	AUC
GODE	0.988414	0.864000	0.900000	0.881633	0.996227
LOF (Breunig et al., 2000)	0.943268	0.412698	0.433333	0.422764	0.819111
kNN (Ramaswamy et al., 2000)	0.987215	0.849206	0.891667	0.869919	0.984438
CBLOF (He et al., 2003)	0.980823	0.785714	0.825000	0.804878	0.970545

OCSVM_Bunny

np.random.seed(77)
clf = OCSVM(nu=0.05)
clf.fit(X)

OCSVM(cache_size=200, coef0=0.0, contamination=0.1, degree=3, gamma='auto',
   kernel='rbf', max_iter=-1, nu=0.05, shrinking=True, tol=0.001,
   verbose=False)

outlier_OSVM_one = list(clf.predict(X))

_conf = Conf_matrx(outlier_true_bunny,outlier_OSVM_one)
_conf.conf("OCSVM (Sch ̈olkopf et al., 2001)")

Accuracy: 0.917
Precision: 0.323
Recall: 0.675
F1 Score: 0.437

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 2.294
Precision(TP / TP + FP): 0.323
Recall(TP / TP + FN): 0.675
F1 Score(2*precision*recall/precision+recall): 0.437

fpr, tpr, thresh = roc_curve(outlier_true_bunny,clf.decision_function(X))

auc(fpr, tpr)

0.8575500069939852

tab_bunny = pd.concat([tab_bunny,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_bunny

	Accuracy	Precision	Recall	F1	AUC
GODE	0.988414	0.864000	0.900000	0.881633	0.996227
LOF (Breunig et al., 2000)	0.943268	0.412698	0.433333	0.422764	0.819111
kNN (Ramaswamy et al., 2000)	0.987215	0.849206	0.891667	0.869919	0.984438
CBLOF (He et al., 2003)	0.980823	0.785714	0.825000	0.804878	0.970545
OCSVM (Sch ̈olkopf et al., 2001)	0.916500	0.322709	0.675000	0.436658	0.857550

MCD_Bunny

clf = MCD(contamination=0.05 , random_state = 77)
clf.fit(_df[['x', 'y','fnoise']])

MCD(assume_centered=False, contamination=0.05, random_state=77,
  store_precision=True, support_fraction=None)

outlier_MCD_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_bunny,outlier_MCD_one)
_conf.conf("MCD (Hardin and Rocke, 2004)")

Accuracy: 0.978
Precision: 0.762
Recall: 0.800
F1 Score: 0.780

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 2.449
Precision(TP / TP + FP): 0.762
Recall(TP / TP + FN): 0.800
F1 Score(2*precision*recall/precision+recall): 0.780

fpr, tpr, thresh = roc_curve(outlier_true_bunny,clf.decision_function(_df[['x', 'y','fnoise']]))

auc(fpr, tpr)

0.9723982375157364

tab_bunny = pd.concat([tab_bunny,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_bunny

	Accuracy	Precision	Recall	F1	AUC
GODE	0.988414	0.864000	0.900000	0.881633	0.996227
LOF (Breunig et al., 2000)	0.943268	0.412698	0.433333	0.422764	0.819111
kNN (Ramaswamy et al., 2000)	0.987215	0.849206	0.891667	0.869919	0.984438
CBLOF (He et al., 2003)	0.980823	0.785714	0.825000	0.804878	0.970545
OCSVM (Sch ̈olkopf et al., 2001)	0.916500	0.322709	0.675000	0.436658	0.857550
MCD (Hardin and Rocke, 2004)	0.978426	0.761905	0.800000	0.780488	0.972398

Feature Bagging_Bunny

clf = FeatureBagging(contamination=0.05, random_state=77)
clf.fit(_df[['x', 'y','fnoise']])

FeatureBagging(base_estimator=None, bootstrap_features=False,
        check_detector=True, check_estimator=False, combination='average',
        contamination=0.05, estimator_params={}, max_features=1.0,
        n_estimators=10, n_jobs=1, random_state=77, verbose=0)

outlier_FeatureBagging_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_bunny,outlier_FeatureBagging_one)
_conf.conf("Feature Bagging (Lazarevic and Kumar, 2005)")

Accuracy: 0.949
Precision: 0.468
Recall: 0.492
F1 Score: 0.480

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 2.375
Precision(TP / TP + FP): 0.468
Recall(TP / TP + FN): 0.492
F1 Score(2*precision*recall/precision+recall): 0.480

fpr, tpr, thresh = roc_curve(outlier_true_bunny,clf.decision_function(_df[['x', 'y','fnoise']]))

auc(fpr, tpr)

0.8338089243250805

tab_bunny = pd.concat([tab_bunny,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_bunny

	Accuracy	Precision	Recall	F1	AUC
GODE	0.988414	0.864000	0.900000	0.881633	0.996227
LOF (Breunig et al., 2000)	0.943268	0.412698	0.433333	0.422764	0.819111
kNN (Ramaswamy et al., 2000)	0.987215	0.849206	0.891667	0.869919	0.984438
CBLOF (He et al., 2003)	0.980823	0.785714	0.825000	0.804878	0.970545
OCSVM (Sch ̈olkopf et al., 2001)	0.916500	0.322709	0.675000	0.436658	0.857550
MCD (Hardin and Rocke, 2004)	0.978426	0.761905	0.800000	0.780488	0.972398
Feature Bagging (Lazarevic and Kumar, 2005)	0.948861	0.468254	0.491667	0.479675	0.833809

ABOD_Bunny

np.random.seed(77)
clf = ABOD(contamination=0.05)
clf.fit(_df[['x', 'y','fnoise']])

ABOD(contamination=0.05, method='fast', n_neighbors=5)

outlier_ABOD_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_bunny,outlier_ABOD_one)
_conf.conf("ABOD (Kriegel et al., 2008)")

Accuracy: 0.979
Precision: 0.770
Recall: 0.808
F1 Score: 0.789

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 2.451
Precision(TP / TP + FP): 0.770
Recall(TP / TP + FN): 0.808
F1 Score(2*precision*recall/precision+recall): 0.789

fpr, tpr, thresh = roc_curve(outlier_true_bunny,clf.decision_function(_df[['x', 'y','fnoise']]))

auc(fpr, tpr)

0.9715764442579382

tab_bunny = pd.concat([tab_bunny,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_bunny

	Accuracy	Precision	Recall	F1	AUC
GODE	0.988414	0.864000	0.900000	0.881633	0.996227
LOF (Breunig et al., 2000)	0.943268	0.412698	0.433333	0.422764	0.819111
kNN (Ramaswamy et al., 2000)	0.987215	0.849206	0.891667	0.869919	0.984438
CBLOF (He et al., 2003)	0.980823	0.785714	0.825000	0.804878	0.970545
OCSVM (Sch ̈olkopf et al., 2001)	0.916500	0.322709	0.675000	0.436658	0.857550
MCD (Hardin and Rocke, 2004)	0.978426	0.761905	0.800000	0.780488	0.972398
Feature Bagging (Lazarevic and Kumar, 2005)	0.948861	0.468254	0.491667	0.479675	0.833809
ABOD (Kriegel et al., 2008)	0.979225	0.769841	0.808333	0.788618	0.971576

IForest_Bunny

clf = IForest(contamination=0.05,random_state=77)
clf.fit(_df[['x', 'y','fnoise']])

IForest(behaviour='old', bootstrap=False, contamination=0.05,
    max_features=1.0, max_samples='auto', n_estimators=100, n_jobs=1,
    random_state=77, verbose=0)

outlier_IForest_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_bunny,outlier_IForest_one)
_conf.conf("Isolation Forest (Liu et al., 2008)")

Accuracy: 0.972
Precision: 0.698
Recall: 0.733
F1 Score: 0.715

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 2.433
Precision(TP / TP + FP): 0.698
Recall(TP / TP + FN): 0.733
F1 Score(2*precision*recall/precision+recall): 0.715

fpr, tpr, thresh = roc_curve(outlier_true_bunny,clf.decision_function(_df[['x', 'y','fnoise']]))

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/base.py:457: UserWarning: X has feature names, but IsolationForest was fitted without feature names
  warnings.warn(

auc(fpr, tpr)

0.967953559938453

tab_bunny = pd.concat([tab_bunny,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_bunny

	Accuracy	Precision	Recall	F1	AUC
GODE	0.988414	0.864000	0.900000	0.881633	0.996227
LOF (Breunig et al., 2000)	0.943268	0.412698	0.433333	0.422764	0.819111
kNN (Ramaswamy et al., 2000)	0.987215	0.849206	0.891667	0.869919	0.984438
CBLOF (He et al., 2003)	0.980823	0.785714	0.825000	0.804878	0.970545
OCSVM (Sch ̈olkopf et al., 2001)	0.916500	0.322709	0.675000	0.436658	0.857550
MCD (Hardin and Rocke, 2004)	0.978426	0.761905	0.800000	0.780488	0.972398
Feature Bagging (Lazarevic and Kumar, 2005)	0.948861	0.468254	0.491667	0.479675	0.833809
ABOD (Kriegel et al., 2008)	0.979225	0.769841	0.808333	0.788618	0.971576
Isolation Forest (Liu et al., 2008)	0.972034	0.698413	0.733333	0.715447	0.967954

HBOS_Bunny

np.random.seed(77)
clf = HBOS(contamination=0.05)
clf.fit(_df[['x', 'y','fnoise']])

HBOS(alpha=0.1, contamination=0.05, n_bins=10, tol=0.5)

outlier_HBOS_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_bunny,outlier_HBOS_one)
_conf.conf("HBOS (Goldstein and Dengel, 2012)")

Accuracy: 0.932
Precision: 0.302
Recall: 0.317
F1 Score: 0.309

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 2.333
Precision(TP / TP + FP): 0.302
Recall(TP / TP + FN): 0.317
F1 Score(2*precision*recall/precision+recall): 0.309

fpr, tpr, thresh = roc_curve(outlier_true_bunny,clf.decision_function(_df[['x', 'y','fnoise']]))

auc(fpr, tpr)

0.8591918450132886

tab_bunny = pd.concat([tab_bunny,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_bunny

	Accuracy	Precision	Recall	F1	AUC
GODE	0.988414	0.864000	0.900000	0.881633	0.996227
LOF (Breunig et al., 2000)	0.943268	0.412698	0.433333	0.422764	0.819111
kNN (Ramaswamy et al., 2000)	0.987215	0.849206	0.891667	0.869919	0.984438
CBLOF (He et al., 2003)	0.980823	0.785714	0.825000	0.804878	0.970545
OCSVM (Sch ̈olkopf et al., 2001)	0.916500	0.322709	0.675000	0.436658	0.857550
MCD (Hardin and Rocke, 2004)	0.978426	0.761905	0.800000	0.780488	0.972398
Feature Bagging (Lazarevic and Kumar, 2005)	0.948861	0.468254	0.491667	0.479675	0.833809
ABOD (Kriegel et al., 2008)	0.979225	0.769841	0.808333	0.788618	0.971576
Isolation Forest (Liu et al., 2008)	0.972034	0.698413	0.733333	0.715447	0.967954
HBOS (Goldstein and Dengel, 2012)	0.932082	0.301587	0.316667	0.308943	0.859192

SOS_Bunny

np.random.seed(77)
clf = SOS(contamination=0.05)
clf.fit(_df[['x', 'y','fnoise']])

SOS(contamination=0.05, eps=1e-05, metric='euclidean', perplexity=4.5)

outlier_SOS_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_bunny,outlier_SOS_one)
_conf.conf("SOS (Janssens et al., 2012)")

Accuracy: 0.909
Precision: 0.071
Recall: 0.075
F1 Score: 0.073

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 2.275
Precision(TP / TP + FP): 0.071
Recall(TP / TP + FN): 0.075
F1 Score(2*precision*recall/precision+recall): 0.073

fpr, tpr, thresh = roc_curve(outlier_true_bunny,clf.decision_function(_df[['x', 'y','fnoise']]))

auc(fpr, tpr)

0.5574311092460484

tab_bunny = pd.concat([tab_bunny,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_bunny

	Accuracy	Precision	Recall	F1	AUC
GODE	0.988414	0.864000	0.900000	0.881633	0.996227
LOF (Breunig et al., 2000)	0.943268	0.412698	0.433333	0.422764	0.819111
kNN (Ramaswamy et al., 2000)	0.987215	0.849206	0.891667	0.869919	0.984438
CBLOF (He et al., 2003)	0.980823	0.785714	0.825000	0.804878	0.970545
OCSVM (Sch ̈olkopf et al., 2001)	0.916500	0.322709	0.675000	0.436658	0.857550
MCD (Hardin and Rocke, 2004)	0.978426	0.761905	0.800000	0.780488	0.972398
Feature Bagging (Lazarevic and Kumar, 2005)	0.948861	0.468254	0.491667	0.479675	0.833809
ABOD (Kriegel et al., 2008)	0.979225	0.769841	0.808333	0.788618	0.971576
Isolation Forest (Liu et al., 2008)	0.972034	0.698413	0.733333	0.715447	0.967954
HBOS (Goldstein and Dengel, 2012)	0.932082	0.301587	0.316667	0.308943	0.859192
SOS (Janssens et al., 2012)	0.908909	0.071429	0.075000	0.073171	0.557431

SO_GAAL_Bunny

np.random.seed(77)
clf = SO_GAAL(contamination=0.05)
clf.fit(_df[['x', 'y','fnoise']])

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/keras/src/optimizers/legacy/gradient_descent.py:114: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.
  super().__init__(name, **kwargs)

Epoch 1 of 60

Testing for epoch 1 index 1:

Testing for epoch 1 index 2:

Testing for epoch 1 index 3:

Testing for epoch 1 index 4:

Testing for epoch 1 index 5:
Epoch 2 of 60

Testing for epoch 2 index 1:

Testing for epoch 2 index 2:

Testing for epoch 2 index 3:

Testing for epoch 2 index 4:

Testing for epoch 2 index 5:
Epoch 3 of 60

Testing for epoch 3 index 1:

Testing for epoch 3 index 2:

Testing for epoch 3 index 3:

Testing for epoch 3 index 4:

Testing for epoch 3 index 5:
Epoch 4 of 60

Testing for epoch 4 index 1:

Testing for epoch 4 index 2:

Testing for epoch 4 index 3:

Testing for epoch 4 index 4:

Testing for epoch 4 index 5:
Epoch 5 of 60

Testing for epoch 5 index 1:

Testing for epoch 5 index 2:

Testing for epoch 5 index 3:

Testing for epoch 5 index 4:

Testing for epoch 5 index 5:
Epoch 6 of 60

Testing for epoch 6 index 1:

Testing for epoch 6 index 2:

Testing for epoch 6 index 3:

Testing for epoch 6 index 4:

Testing for epoch 6 index 5:
Epoch 7 of 60

Testing for epoch 7 index 1:

Testing for epoch 7 index 2:

Testing for epoch 7 index 3:

Testing for epoch 7 index 4:

Testing for epoch 7 index 5:
Epoch 8 of 60

Testing for epoch 8 index 1:

Testing for epoch 8 index 2:

Testing for epoch 8 index 3:

Testing for epoch 8 index 4:

Testing for epoch 8 index 5:
Epoch 9 of 60

Testing for epoch 9 index 1:

Testing for epoch 9 index 2:

Testing for epoch 9 index 3:

Testing for epoch 9 index 4:

Testing for epoch 9 index 5:
Epoch 10 of 60

Testing for epoch 10 index 1:

Testing for epoch 10 index 2:

Testing for epoch 10 index 3:

Testing for epoch 10 index 4:

Testing for epoch 10 index 5:
Epoch 11 of 60

Testing for epoch 11 index 1:

Testing for epoch 11 index 2:

Testing for epoch 11 index 3:

Testing for epoch 11 index 4:

Testing for epoch 11 index 5:
Epoch 12 of 60

Testing for epoch 12 index 1:

Testing for epoch 12 index 2:

Testing for epoch 12 index 3:

Testing for epoch 12 index 4:

Testing for epoch 12 index 5:
Epoch 13 of 60

Testing for epoch 13 index 1:

Testing for epoch 13 index 2:

Testing for epoch 13 index 3:

Testing for epoch 13 index 4:

Testing for epoch 13 index 5:
Epoch 14 of 60

Testing for epoch 14 index 1:

Testing for epoch 14 index 2:

Testing for epoch 14 index 3:

Testing for epoch 14 index 4:

Testing for epoch 14 index 5:
Epoch 15 of 60

Testing for epoch 15 index 1:

Testing for epoch 15 index 2:

Testing for epoch 15 index 3:

Testing for epoch 15 index 4:

Testing for epoch 15 index 5:
Epoch 16 of 60

Testing for epoch 16 index 1:

Testing for epoch 16 index 2:

Testing for epoch 16 index 3:

Testing for epoch 16 index 4:

Testing for epoch 16 index 5:
Epoch 17 of 60

Testing for epoch 17 index 1:

Testing for epoch 17 index 2:

Testing for epoch 17 index 3:

Testing for epoch 17 index 4:

Testing for epoch 17 index 5:
Epoch 18 of 60

Testing for epoch 18 index 1:

Testing for epoch 18 index 2:

Testing for epoch 18 index 3:

Testing for epoch 18 index 4:

Testing for epoch 18 index 5:
Epoch 19 of 60

Testing for epoch 19 index 1:

Testing for epoch 19 index 2:

Testing for epoch 19 index 3:

Testing for epoch 19 index 4:

Testing for epoch 19 index 5:
Epoch 20 of 60

Testing for epoch 20 index 1:

Testing for epoch 20 index 2:

Testing for epoch 20 index 3:

Testing for epoch 20 index 4:

Testing for epoch 20 index 5:
Epoch 21 of 60

Testing for epoch 21 index 1:

Testing for epoch 21 index 2:

Testing for epoch 21 index 3:

Testing for epoch 21 index 4:

Testing for epoch 21 index 5:
Epoch 22 of 60

Testing for epoch 22 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.8892

Testing for epoch 22 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 1.8773

Testing for epoch 22 index 3:
16/16 [==============================] - 0s 968us/step - loss: 1.8953

Testing for epoch 22 index 4:
16/16 [==============================] - 0s 1ms/step - loss: 1.8628

Testing for epoch 22 index 5:
16/16 [==============================] - 0s 841us/step - loss: 1.9471
Epoch 23 of 60

Testing for epoch 23 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 1.9137

Testing for epoch 23 index 2:
16/16 [==============================] - 0s 854us/step - loss: 1.9870

Testing for epoch 23 index 3:
16/16 [==============================] - 0s 2ms/step - loss: 1.9878

Testing for epoch 23 index 4:
16/16 [==============================] - 0s 1ms/step - loss: 1.9311

Testing for epoch 23 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 1.9295
Epoch 24 of 60

Testing for epoch 24 index 1:
16/16 [==============================] - 0s 888us/step - loss: 1.9993

Testing for epoch 24 index 2:
16/16 [==============================] - 0s 807us/step - loss: 1.9489

Testing for epoch 24 index 3:
16/16 [==============================] - 0s 1ms/step - loss: 1.9849

Testing for epoch 24 index 4:
16/16 [==============================] - 0s 813us/step - loss: 1.9797

Testing for epoch 24 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 1.9702
Epoch 25 of 60

Testing for epoch 25 index 1:
16/16 [==============================] - 0s 869us/step - loss: 2.0650

Testing for epoch 25 index 2:
16/16 [==============================] - 0s 798us/step - loss: 2.0078

Testing for epoch 25 index 3:
16/16 [==============================] - 0s 1ms/step - loss: 2.0171

Testing for epoch 25 index 4:
16/16 [==============================] - 0s 922us/step - loss: 2.0466

Testing for epoch 25 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 2.0635
Epoch 26 of 60

Testing for epoch 26 index 1:
16/16 [==============================] - 0s 956us/step - loss: 2.0758

Testing for epoch 26 index 2:
16/16 [==============================] - 0s 834us/step - loss: 2.0865

Testing for epoch 26 index 3:
16/16 [==============================] - 0s 2ms/step - loss: 2.0039

Testing for epoch 26 index 4:
16/16 [==============================] - 0s 2ms/step - loss: 2.0798

Testing for epoch 26 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 2.0561
Epoch 27 of 60

Testing for epoch 27 index 1:
16/16 [==============================] - 0s 950us/step - loss: 2.1265

Testing for epoch 27 index 2:
16/16 [==============================] - 0s 801us/step - loss: 2.0775

Testing for epoch 27 index 3:
16/16 [==============================] - 0s 792us/step - loss: 2.0951

Testing for epoch 27 index 4:
16/16 [==============================] - 0s 805us/step - loss: 2.1185

Testing for epoch 27 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 2.1396
Epoch 28 of 60

Testing for epoch 28 index 1:
16/16 [==============================] - 0s 807us/step - loss: 2.1326

Testing for epoch 28 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.1569

Testing for epoch 28 index 3:
16/16 [==============================] - 0s 849us/step - loss: 2.0954

Testing for epoch 28 index 4:
16/16 [==============================] - 0s 2ms/step - loss: 2.0795

Testing for epoch 28 index 5:
16/16 [==============================] - 0s 1ms/step - loss: 2.1101
Epoch 29 of 60

Testing for epoch 29 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 2.1282

Testing for epoch 29 index 2:
16/16 [==============================] - 0s 830us/step - loss: 2.0985

Testing for epoch 29 index 3:
16/16 [==============================] - 0s 2ms/step - loss: 2.0946

Testing for epoch 29 index 4:
16/16 [==============================] - 0s 1ms/step - loss: 2.0580

Testing for epoch 29 index 5:
16/16 [==============================] - 0s 1ms/step - loss: 2.0928
Epoch 30 of 60

Testing for epoch 30 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.1236

Testing for epoch 30 index 2:
16/16 [==============================] - 0s 800us/step - loss: 2.1554

Testing for epoch 30 index 3:
16/16 [==============================] - 0s 1ms/step - loss: 2.1987

Testing for epoch 30 index 4:
16/16 [==============================] - 0s 793us/step - loss: 2.1537

Testing for epoch 30 index 5:
16/16 [==============================] - 0s 870us/step - loss: 2.1312
Epoch 31 of 60

Testing for epoch 31 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 2.1679

Testing for epoch 31 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.2280

Testing for epoch 31 index 3:
16/16 [==============================] - 0s 834us/step - loss: 2.1869

Testing for epoch 31 index 4:
16/16 [==============================] - 0s 808us/step - loss: 2.1695

Testing for epoch 31 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 2.2448
Epoch 32 of 60

Testing for epoch 32 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.1599

Testing for epoch 32 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 2.2401

Testing for epoch 32 index 3:
16/16 [==============================] - 0s 755us/step - loss: 2.2098

Testing for epoch 32 index 4:
16/16 [==============================] - 0s 840us/step - loss: 2.2194

Testing for epoch 32 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 2.2362
Epoch 33 of 60

Testing for epoch 33 index 1:
16/16 [==============================] - 0s 793us/step - loss: 2.2603

Testing for epoch 33 index 2:
16/16 [==============================] - 0s 939us/step - loss: 2.1833

Testing for epoch 33 index 3:
16/16 [==============================] - 0s 2ms/step - loss: 2.2989

Testing for epoch 33 index 4:
16/16 [==============================] - 0s 823us/step - loss: 2.2219

Testing for epoch 33 index 5:
16/16 [==============================] - 0s 801us/step - loss: 2.2803
Epoch 34 of 60

Testing for epoch 34 index 1:
16/16 [==============================] - 0s 784us/step - loss: 2.2164

Testing for epoch 34 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.2358

Testing for epoch 34 index 3:
16/16 [==============================] - 0s 2ms/step - loss: 2.1974

Testing for epoch 34 index 4:
16/16 [==============================] - 0s 889us/step - loss: 2.2659

Testing for epoch 34 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 2.2727
Epoch 35 of 60

Testing for epoch 35 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.2802

Testing for epoch 35 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.2867

Testing for epoch 35 index 3:
16/16 [==============================] - 0s 1ms/step - loss: 2.2722

Testing for epoch 35 index 4:
16/16 [==============================] - 0s 857us/step - loss: 2.2683

Testing for epoch 35 index 5:
16/16 [==============================] - 0s 883us/step - loss: 2.3110
Epoch 36 of 60

Testing for epoch 36 index 1:
16/16 [==============================] - 0s 781us/step - loss: 2.2822

Testing for epoch 36 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.3013

Testing for epoch 36 index 3:
16/16 [==============================] - 0s 2ms/step - loss: 2.3260

Testing for epoch 36 index 4:
16/16 [==============================] - 0s 911us/step - loss: 2.3637

Testing for epoch 36 index 5:
16/16 [==============================] - 0s 814us/step - loss: 2.2913
Epoch 37 of 60

Testing for epoch 37 index 1:
16/16 [==============================] - 0s 814us/step - loss: 2.3133

Testing for epoch 37 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 2.2731

Testing for epoch 37 index 3:
16/16 [==============================] - 0s 971us/step - loss: 2.3091

Testing for epoch 37 index 4:
16/16 [==============================] - 0s 2ms/step - loss: 2.3101

Testing for epoch 37 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 2.2987
Epoch 38 of 60

Testing for epoch 38 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 2.3002

Testing for epoch 38 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 2.2867

Testing for epoch 38 index 3:
16/16 [==============================] - 0s 843us/step - loss: 2.2980

Testing for epoch 38 index 4:
16/16 [==============================] - 0s 806us/step - loss: 2.3380

Testing for epoch 38 index 5:
16/16 [==============================] - 0s 1ms/step - loss: 2.2815
Epoch 39 of 60

Testing for epoch 39 index 1:
16/16 [==============================] - 0s 1ms/step - loss: 2.3563

Testing for epoch 39 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.3031

Testing for epoch 39 index 3:
16/16 [==============================] - 0s 868us/step - loss: 2.3611

Testing for epoch 39 index 4:
16/16 [==============================] - 0s 1ms/step - loss: 2.3812

Testing for epoch 39 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 2.3761
Epoch 40 of 60

Testing for epoch 40 index 1:
16/16 [==============================] - 0s 838us/step - loss: 2.3955

Testing for epoch 40 index 2:
16/16 [==============================] - 0s 796us/step - loss: 2.4065

Testing for epoch 40 index 3:
16/16 [==============================] - 0s 1ms/step - loss: 2.4450

Testing for epoch 40 index 4:
16/16 [==============================] - 0s 840us/step - loss: 2.3389

Testing for epoch 40 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 2.3475
Epoch 41 of 60

Testing for epoch 41 index 1:
16/16 [==============================] - 0s 821us/step - loss: 2.3851

Testing for epoch 41 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.3873

Testing for epoch 41 index 3:
16/16 [==============================] - 0s 2ms/step - loss: 2.3659

Testing for epoch 41 index 4:
16/16 [==============================] - 0s 2ms/step - loss: 2.3792

Testing for epoch 41 index 5:
16/16 [==============================] - 0s 1ms/step - loss: 2.3896
Epoch 42 of 60

Testing for epoch 42 index 1:
16/16 [==============================] - 0s 985us/step - loss: 2.4595

Testing for epoch 42 index 2:
16/16 [==============================] - 0s 821us/step - loss: 2.4335

Testing for epoch 42 index 3:
16/16 [==============================] - 0s 1ms/step - loss: 2.3803

Testing for epoch 42 index 4:
16/16 [==============================] - 0s 827us/step - loss: 2.3513

Testing for epoch 42 index 5:
16/16 [==============================] - 0s 809us/step - loss: 2.4367
Epoch 43 of 60

Testing for epoch 43 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.3864

Testing for epoch 43 index 2:
16/16 [==============================] - 0s 795us/step - loss: 2.3912

Testing for epoch 43 index 3:
16/16 [==============================] - 0s 2ms/step - loss: 2.4194

Testing for epoch 43 index 4:
16/16 [==============================] - 0s 799us/step - loss: 2.5205

Testing for epoch 43 index 5:
16/16 [==============================] - 0s 815us/step - loss: 2.4733
Epoch 44 of 60

Testing for epoch 44 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.4465

Testing for epoch 44 index 2:
16/16 [==============================] - 0s 1ms/step - loss: 2.4298

Testing for epoch 44 index 3:
16/16 [==============================] - 0s 2ms/step - loss: 2.4497

Testing for epoch 44 index 4:
16/16 [==============================] - 0s 821us/step - loss: 2.4719

Testing for epoch 44 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 2.4463
Epoch 45 of 60

Testing for epoch 45 index 1:
16/16 [==============================] - 0s 824us/step - loss: 2.5313

Testing for epoch 45 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.4766

Testing for epoch 45 index 3:
16/16 [==============================] - 0s 956us/step - loss: 2.4820

Testing for epoch 45 index 4:
16/16 [==============================] - 0s 2ms/step - loss: 2.4411

Testing for epoch 45 index 5:
16/16 [==============================] - 0s 813us/step - loss: 2.4633
Epoch 46 of 60

Testing for epoch 46 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.4844

Testing for epoch 46 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.4777

Testing for epoch 46 index 3:
16/16 [==============================] - 0s 826us/step - loss: 2.4881

Testing for epoch 46 index 4:
16/16 [==============================] - 0s 800us/step - loss: 2.5910

Testing for epoch 46 index 5:
16/16 [==============================] - 0s 810us/step - loss: 2.3987
Epoch 47 of 60

Testing for epoch 47 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.5073

Testing for epoch 47 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.4767

Testing for epoch 47 index 3:
16/16 [==============================] - 0s 1ms/step - loss: 2.5076

Testing for epoch 47 index 4:
16/16 [==============================] - 0s 820us/step - loss: 2.4839

Testing for epoch 47 index 5:
16/16 [==============================] - 0s 822us/step - loss: 2.4424
Epoch 48 of 60

Testing for epoch 48 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.4959

Testing for epoch 48 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.5135

Testing for epoch 48 index 3:
16/16 [==============================] - 0s 2ms/step - loss: 2.4452

Testing for epoch 48 index 4:
16/16 [==============================] - 0s 2ms/step - loss: 2.5420

Testing for epoch 48 index 5:
16/16 [==============================] - 0s 812us/step - loss: 2.5427
Epoch 49 of 60

Testing for epoch 49 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.4873

Testing for epoch 49 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.5430

Testing for epoch 49 index 3:
16/16 [==============================] - 0s 838us/step - loss: 2.4718

Testing for epoch 49 index 4:
16/16 [==============================] - 0s 2ms/step - loss: 2.5524

Testing for epoch 49 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 2.4794
Epoch 50 of 60

Testing for epoch 50 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.5831

Testing for epoch 50 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.5607

Testing for epoch 50 index 3:
16/16 [==============================] - 0s 835us/step - loss: 2.5869

Testing for epoch 50 index 4:
16/16 [==============================] - 0s 820us/step - loss: 2.5292

Testing for epoch 50 index 5:
16/16 [==============================] - 0s 818us/step - loss: 2.5355
Epoch 51 of 60

Testing for epoch 51 index 1:
16/16 [==============================] - 0s 794us/step - loss: 2.6074

Testing for epoch 51 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.5498

Testing for epoch 51 index 3:
16/16 [==============================] - 0s 798us/step - loss: 2.5585

Testing for epoch 51 index 4:
16/16 [==============================] - 0s 1ms/step - loss: 2.5851

Testing for epoch 51 index 5:
16/16 [==============================] - 0s 810us/step - loss: 2.5378
Epoch 52 of 60

Testing for epoch 52 index 1:
16/16 [==============================] - 0s 805us/step - loss: 2.5224

Testing for epoch 52 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.5417

Testing for epoch 52 index 3:
16/16 [==============================] - 0s 2ms/step - loss: 2.5603

Testing for epoch 52 index 4:
16/16 [==============================] - 0s 2ms/step - loss: 2.5464

Testing for epoch 52 index 5:
16/16 [==============================] - 0s 1ms/step - loss: 2.5803
Epoch 53 of 60

Testing for epoch 53 index 1:
16/16 [==============================] - 0s 819us/step - loss: 2.5619

Testing for epoch 53 index 2:
16/16 [==============================] - 0s 797us/step - loss: 2.5412

Testing for epoch 53 index 3:
16/16 [==============================] - 0s 1ms/step - loss: 2.6601

Testing for epoch 53 index 4:
16/16 [==============================] - 0s 813us/step - loss: 2.4935

Testing for epoch 53 index 5:
16/16 [==============================] - 0s 1ms/step - loss: 2.5908
Epoch 54 of 60

Testing for epoch 54 index 1:
16/16 [==============================] - 0s 839us/step - loss: 2.5420

Testing for epoch 54 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.6080

Testing for epoch 54 index 3:
16/16 [==============================] - 0s 935us/step - loss: 2.5973

Testing for epoch 54 index 4:
16/16 [==============================] - 0s 820us/step - loss: 2.6289

Testing for epoch 54 index 5:
16/16 [==============================] - 0s 811us/step - loss: 2.6870
Epoch 55 of 60

Testing for epoch 55 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.5925

Testing for epoch 55 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.6485

Testing for epoch 55 index 3:
16/16 [==============================] - 0s 2ms/step - loss: 2.6448

Testing for epoch 55 index 4:
16/16 [==============================] - 0s 1ms/step - loss: 2.5929

Testing for epoch 55 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 2.5843
Epoch 56 of 60

Testing for epoch 56 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.6255

Testing for epoch 56 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.6109

Testing for epoch 56 index 3:
16/16 [==============================] - 0s 2ms/step - loss: 2.5776

Testing for epoch 56 index 4:
16/16 [==============================] - 0s 1ms/step - loss: 2.5939

Testing for epoch 56 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 2.6101
Epoch 57 of 60

Testing for epoch 57 index 1:
16/16 [==============================] - 0s 869us/step - loss: 2.6185

Testing for epoch 57 index 2:
16/16 [==============================] - 0s 811us/step - loss: 2.5944

Testing for epoch 57 index 3:
16/16 [==============================] - 0s 845us/step - loss: 2.6853

Testing for epoch 57 index 4:
16/16 [==============================] - 0s 808us/step - loss: 2.6887

Testing for epoch 57 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 2.5760
Epoch 58 of 60

Testing for epoch 58 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.7048

Testing for epoch 58 index 2:
16/16 [==============================] - 0s 939us/step - loss: 2.6857

Testing for epoch 58 index 3:
16/16 [==============================] - 0s 2ms/step - loss: 2.7510

Testing for epoch 58 index 4:
16/16 [==============================] - 0s 2ms/step - loss: 2.6191

Testing for epoch 58 index 5:
16/16 [==============================] - 0s 812us/step - loss: 2.6330
Epoch 59 of 60

Testing for epoch 59 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.6354

Testing for epoch 59 index 2:
16/16 [==============================] - 0s 806us/step - loss: 2.6133

Testing for epoch 59 index 3:
16/16 [==============================] - 0s 895us/step - loss: 2.6335

Testing for epoch 59 index 4:
16/16 [==============================] - 0s 2ms/step - loss: 2.6475

Testing for epoch 59 index 5:
16/16 [==============================] - 0s 2ms/step - loss: 2.6326
Epoch 60 of 60

Testing for epoch 60 index 1:
16/16 [==============================] - 0s 2ms/step - loss: 2.7069

Testing for epoch 60 index 2:
16/16 [==============================] - 0s 2ms/step - loss: 2.7578

Testing for epoch 60 index 3:
16/16 [==============================] - 0s 2ms/step - loss: 2.7084

Testing for epoch 60 index 4:
16/16 [==============================] - 0s 2ms/step - loss: 2.6440

Testing for epoch 60 index 5:
16/16 [==============================] - 0s 914us/step - loss: 2.6770
79/79 [==============================] - 0s 732us/step

SO_GAAL(contamination=0.05, lr_d=0.01, lr_g=0.0001, momentum=0.9,
    stop_epochs=20)

outlier_SO_GAAL_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_bunny,outlier_SO_GAAL_one)
_conf.conf("SO-GAAL (Liu et al., 2019)")

Accuracy: 0.952
Precision: 0.000
Recall: 0.000
F1 Score: 0.000

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/metrics/_classification.py:1469: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.
  _warn_prf(average, modifier, msg_start, len(result))

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 2.383
Precision(TP / TP + FP): nan
Recall(TP / TP + FN): 0.000
F1 Score(2*precision*recall/precision+recall): nan

/tmp/ipykernel_3852735/4166638268.py:3: RuntimeWarning: invalid value encountered in long_scalars
  print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
/tmp/ipykernel_3852735/4166638268.py:5: RuntimeWarning: invalid value encountered in long_scalars
  print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

fpr, tpr, thresh = roc_curve(outlier_true_bunny,clf.decision_function(_df[['x', 'y','fnoise']]))

79/79 [==============================] - 0s 527us/step

auc(fpr, tpr)

0.6730102112183523

tab_bunny = pd.concat([tab_bunny,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_bunny

	Accuracy	Precision	Recall	F1	AUC
GODE	0.988414	0.864000	0.900000	0.881633	0.996227
LOF (Breunig et al., 2000)	0.943268	0.412698	0.433333	0.422764	0.819111
kNN (Ramaswamy et al., 2000)	0.987215	0.849206	0.891667	0.869919	0.984438
CBLOF (He et al., 2003)	0.980823	0.785714	0.825000	0.804878	0.970545
OCSVM (Sch ̈olkopf et al., 2001)	0.916500	0.322709	0.675000	0.436658	0.857550
MCD (Hardin and Rocke, 2004)	0.978426	0.761905	0.800000	0.780488	0.972398
Feature Bagging (Lazarevic and Kumar, 2005)	0.948861	0.468254	0.491667	0.479675	0.833809
ABOD (Kriegel et al., 2008)	0.979225	0.769841	0.808333	0.788618	0.971576
Isolation Forest (Liu et al., 2008)	0.972034	0.698413	0.733333	0.715447	0.967954
HBOS (Goldstein and Dengel, 2012)	0.932082	0.301587	0.316667	0.308943	0.859192
SOS (Janssens et al., 2012)	0.908909	0.071429	0.075000	0.073171	0.557431
SO-GAAL (Liu et al., 2019)	0.952058	0.000000	0.000000	0.000000	0.673010

MO_GAAL_Bunny

np.random.seed(77)
clf = MO_GAAL(contamination=0.05)
clf.fit(_df[['x', 'y','fnoise']])

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/keras/src/optimizers/legacy/gradient_descent.py:114: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.
  super().__init__(name, **kwargs)

Epoch 1 of 60

Testing for epoch 1 index 1:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 1 index 2:
79/79 [==============================] - 0s 2ms/step

Testing for epoch 1 index 3:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 1 index 4:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 1 index 5:
79/79 [==============================] - 0s 822us/step
Epoch 2 of 60

Testing for epoch 2 index 1:
79/79 [==============================] - 0s 556us/step

Testing for epoch 2 index 2:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 2 index 3:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 2 index 4:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 2 index 5:
79/79 [==============================] - 0s 1ms/step
Epoch 3 of 60

Testing for epoch 3 index 1:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 3 index 2:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 3 index 3:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 3 index 4:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 3 index 5:
79/79 [==============================] - 0s 1ms/step
Epoch 4 of 60

Testing for epoch 4 index 1:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 4 index 2:
79/79 [==============================] - 0s 832us/step

Testing for epoch 4 index 3:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 4 index 4:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 4 index 5:
79/79 [==============================] - 0s 2ms/step
Epoch 5 of 60

Testing for epoch 5 index 1:
79/79 [==============================] - 0s 540us/step

Testing for epoch 5 index 2:
79/79 [==============================] - 0s 539us/step

Testing for epoch 5 index 3:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 5 index 4:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 5 index 5:
79/79 [==============================] - 0s 792us/step
Epoch 6 of 60

Testing for epoch 6 index 1:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 6 index 2:
79/79 [==============================] - 0s 527us/step

Testing for epoch 6 index 3:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 6 index 4:
79/79 [==============================] - 0s 928us/step

Testing for epoch 6 index 5:
79/79 [==============================] - 0s 1ms/step
Epoch 7 of 60

Testing for epoch 7 index 1:
79/79 [==============================] - 0s 2ms/step

Testing for epoch 7 index 2:
79/79 [==============================] - 0s 994us/step

Testing for epoch 7 index 3:
79/79 [==============================] - 0s 2ms/step

Testing for epoch 7 index 4:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 7 index 5:
79/79 [==============================] - 0s 612us/step
Epoch 8 of 60

Testing for epoch 8 index 1:
79/79 [==============================] - 0s 919us/step

Testing for epoch 8 index 2:
79/79 [==============================] - 0s 990us/step

Testing for epoch 8 index 3:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 8 index 4:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 8 index 5:
79/79 [==============================] - 0s 901us/step
Epoch 9 of 60

Testing for epoch 9 index 1:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 9 index 2:
79/79 [==============================] - 0s 567us/step

Testing for epoch 9 index 3:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 9 index 4:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 9 index 5:
79/79 [==============================] - 0s 568us/step
Epoch 10 of 60

Testing for epoch 10 index 1:
79/79 [==============================] - 0s 6ms/step

Testing for epoch 10 index 2:
79/79 [==============================] - 0s 534us/step

Testing for epoch 10 index 3:
79/79 [==============================] - 0s 541us/step

Testing for epoch 10 index 4:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 10 index 5:
79/79 [==============================] - 0s 527us/step
Epoch 11 of 60

Testing for epoch 11 index 1:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 11 index 2:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 11 index 3:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 11 index 4:
79/79 [==============================] - 0s 999us/step

Testing for epoch 11 index 5:
79/79 [==============================] - 0s 2ms/step
Epoch 12 of 60

Testing for epoch 12 index 1:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 12 index 2:
79/79 [==============================] - 0s 2ms/step

Testing for epoch 12 index 3:
79/79 [==============================] - 0s 927us/step

Testing for epoch 12 index 4:
79/79 [==============================] - 0s 2ms/step

Testing for epoch 12 index 5:
79/79 [==============================] - 0s 544us/step
Epoch 13 of 60

Testing for epoch 13 index 1:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 13 index 2:
79/79 [==============================] - 0s 999us/step

Testing for epoch 13 index 3:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 13 index 4:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 13 index 5:
79/79 [==============================] - 0s 538us/step
Epoch 14 of 60

Testing for epoch 14 index 1:
79/79 [==============================] - 0s 533us/step

Testing for epoch 14 index 2:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 14 index 3:
79/79 [==============================] - 0s 521us/step

Testing for epoch 14 index 4:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 14 index 5:
79/79 [==============================] - 0s 535us/step
Epoch 15 of 60

Testing for epoch 15 index 1:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 15 index 2:
79/79 [==============================] - 0s 2ms/step

Testing for epoch 15 index 3:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 15 index 4:
79/79 [==============================] - 0s 534us/step

Testing for epoch 15 index 5:
79/79 [==============================] - 0s 1ms/step
Epoch 16 of 60

Testing for epoch 16 index 1:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 16 index 2:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 16 index 3:
79/79 [==============================] - 0s 2ms/step

Testing for epoch 16 index 4:
79/79 [==============================] - 0s 2ms/step

Testing for epoch 16 index 5:
79/79 [==============================] - 0s 533us/step
Epoch 17 of 60

Testing for epoch 17 index 1:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 17 index 2:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 17 index 3:
79/79 [==============================] - 0s 532us/step

Testing for epoch 17 index 4:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 17 index 5:
79/79 [==============================] - 0s 812us/step
Epoch 18 of 60

Testing for epoch 18 index 1:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 18 index 2:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 18 index 3:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 18 index 4:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 18 index 5:
79/79 [==============================] - 0s 1ms/step
Epoch 19 of 60

Testing for epoch 19 index 1:
79/79 [==============================] - 0s 989us/step

Testing for epoch 19 index 2:
79/79 [==============================] - 0s 531us/step

Testing for epoch 19 index 3:
79/79 [==============================] - 0s 969us/step

Testing for epoch 19 index 4:
79/79 [==============================] - 0s 2ms/step

Testing for epoch 19 index 5:
79/79 [==============================] - 0s 1ms/step
Epoch 20 of 60

Testing for epoch 20 index 1:
79/79 [==============================] - 0s 672us/step

Testing for epoch 20 index 2:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 20 index 3:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 20 index 4:
79/79 [==============================] - 0s 965us/step

Testing for epoch 20 index 5:
79/79 [==============================] - 0s 2ms/step
Epoch 21 of 60

Testing for epoch 21 index 1:
79/79 [==============================] - 0s 1ms/step

Testing for epoch 21 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1840
16/16 [==============================] - 0s 861us/step - loss: 1.4152
16/16 [==============================] - 0s 2ms/step - loss: 1.6292
16/16 [==============================] - 0s 2ms/step - loss: 1.6820
16/16 [==============================] - 0s 2ms/step - loss: 1.6830
16/16 [==============================] - 0s 863us/step - loss: 1.6735
16/16 [==============================] - 0s 2ms/step - loss: 1.6625
16/16 [==============================] - 0s 2ms/step - loss: 1.6544
16/16 [==============================] - 0s 2ms/step - loss: 1.6507
16/16 [==============================] - 0s 2ms/step - loss: 1.6493

Testing for epoch 21 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.1812
16/16 [==============================] - 0s 900us/step - loss: 1.4503
16/16 [==============================] - 0s 853us/step - loss: 1.6832
16/16 [==============================] - 0s 846us/step - loss: 1.7404
16/16 [==============================] - 0s 1ms/step - loss: 1.7416
16/16 [==============================] - 0s 848us/step - loss: 1.7307
16/16 [==============================] - 0s 859us/step - loss: 1.7183
16/16 [==============================] - 0s 869us/step - loss: 1.7096
16/16 [==============================] - 0s 857us/step - loss: 1.7057
16/16 [==============================] - 0s 864us/step - loss: 1.7044

Testing for epoch 21 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1766
16/16 [==============================] - 0s 915us/step - loss: 1.4131
16/16 [==============================] - 0s 2ms/step - loss: 1.6342
16/16 [==============================] - 0s 890us/step - loss: 1.6873
16/16 [==============================] - 0s 1ms/step - loss: 1.6871
16/16 [==============================] - 0s 1ms/step - loss: 1.6767
16/16 [==============================] - 0s 1ms/step - loss: 1.6644
16/16 [==============================] - 0s 2ms/step - loss: 1.6559
16/16 [==============================] - 0s 2ms/step - loss: 1.6521
16/16 [==============================] - 0s 1ms/step - loss: 1.6509

Testing for epoch 21 index 5:
79/79 [==============================] - 0s 927us/step
16/16 [==============================] - 0s 1ms/step - loss: 0.1745
16/16 [==============================] - 0s 2ms/step - loss: 1.4735
16/16 [==============================] - 0s 2ms/step - loss: 1.7117
16/16 [==============================] - 0s 1ms/step - loss: 1.7696
16/16 [==============================] - 0s 864us/step - loss: 1.7706
16/16 [==============================] - 0s 2ms/step - loss: 1.7616
16/16 [==============================] - 0s 2ms/step - loss: 1.7494
16/16 [==============================] - 0s 2ms/step - loss: 1.7407
16/16 [==============================] - 0s 889us/step - loss: 1.7368
16/16 [==============================] - 0s 2ms/step - loss: 1.7354
Epoch 22 of 60

Testing for epoch 22 index 1:
79/79 [==============================] - 0s 678us/step
16/16 [==============================] - 0s 1ms/step - loss: 0.1723
16/16 [==============================] - 0s 2ms/step - loss: 1.4294
16/16 [==============================] - 0s 901us/step - loss: 1.6597
16/16 [==============================] - 0s 2ms/step - loss: 1.7140
16/16 [==============================] - 0s 833us/step - loss: 1.7139
16/16 [==============================] - 0s 833us/step - loss: 1.7050
16/16 [==============================] - 0s 2ms/step - loss: 1.6925
16/16 [==============================] - 0s 2ms/step - loss: 1.6839
16/16 [==============================] - 0s 1ms/step - loss: 1.6801
16/16 [==============================] - 0s 1ms/step - loss: 1.6788

Testing for epoch 22 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1642
16/16 [==============================] - 0s 937us/step - loss: 1.5060
16/16 [==============================] - 0s 1ms/step - loss: 1.7582
16/16 [==============================] - 0s 2ms/step - loss: 1.8145
16/16 [==============================] - 0s 1ms/step - loss: 1.8126
16/16 [==============================] - 0s 2ms/step - loss: 1.8008
16/16 [==============================] - 0s 1ms/step - loss: 1.7866
16/16 [==============================] - 0s 1ms/step - loss: 1.7772
16/16 [==============================] - 0s 2ms/step - loss: 1.7731
16/16 [==============================] - 0s 2ms/step - loss: 1.7717

Testing for epoch 22 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1661
16/16 [==============================] - 0s 860us/step - loss: 1.4501
16/16 [==============================] - 0s 2ms/step - loss: 1.6971
16/16 [==============================] - 0s 925us/step - loss: 1.7575
16/16 [==============================] - 0s 2ms/step - loss: 1.7591
16/16 [==============================] - 0s 2ms/step - loss: 1.7501
16/16 [==============================] - 0s 2ms/step - loss: 1.7368
16/16 [==============================] - 0s 969us/step - loss: 1.7276
16/16 [==============================] - 0s 2ms/step - loss: 1.7235
16/16 [==============================] - 0s 2ms/step - loss: 1.7221

Testing for epoch 22 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.1628
16/16 [==============================] - 0s 2ms/step - loss: 1.4725
16/16 [==============================] - 0s 2ms/step - loss: 1.7223
16/16 [==============================] - 0s 869us/step - loss: 1.7835
16/16 [==============================] - 0s 2ms/step - loss: 1.7839
16/16 [==============================] - 0s 2ms/step - loss: 1.7738
16/16 [==============================] - 0s 2ms/step - loss: 1.7600
16/16 [==============================] - 0s 886us/step - loss: 1.7510
16/16 [==============================] - 0s 2ms/step - loss: 1.7470
16/16 [==============================] - 0s 899us/step - loss: 1.7457

Testing for epoch 22 index 5:
79/79 [==============================] - 0s 2ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1607
16/16 [==============================] - 0s 2ms/step - loss: 1.4542
16/16 [==============================] - 0s 995us/step - loss: 1.7067
16/16 [==============================] - 0s 830us/step - loss: 1.7676
16/16 [==============================] - 0s 804us/step - loss: 1.7679
16/16 [==============================] - 0s 802us/step - loss: 1.7587
16/16 [==============================] - 0s 827us/step - loss: 1.7455
16/16 [==============================] - 0s 2ms/step - loss: 1.7365
16/16 [==============================] - 0s 811us/step - loss: 1.7324
16/16 [==============================] - 0s 790us/step - loss: 1.7309
Epoch 23 of 60

Testing for epoch 23 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 893us/step - loss: 0.1558
16/16 [==============================] - 0s 2ms/step - loss: 1.4940
16/16 [==============================] - 0s 2ms/step - loss: 1.7608
16/16 [==============================] - 0s 2ms/step - loss: 1.8229
16/16 [==============================] - 0s 2ms/step - loss: 1.8213
16/16 [==============================] - 0s 857us/step - loss: 1.8105
16/16 [==============================] - 0s 792us/step - loss: 1.7952
16/16 [==============================] - 0s 2ms/step - loss: 1.7854
16/16 [==============================] - 0s 2ms/step - loss: 1.7812
16/16 [==============================] - 0s 1ms/step - loss: 1.7798

Testing for epoch 23 index 2:
79/79 [==============================] - 0s 529us/step
16/16 [==============================] - 0s 780us/step - loss: 0.1554
16/16 [==============================] - 0s 806us/step - loss: 1.4708
16/16 [==============================] - 0s 777us/step - loss: 1.7371
16/16 [==============================] - 0s 2ms/step - loss: 1.8014
16/16 [==============================] - 0s 1ms/step - loss: 1.8023
16/16 [==============================] - 0s 821us/step - loss: 1.7935
16/16 [==============================] - 0s 841us/step - loss: 1.7807
16/16 [==============================] - 0s 803us/step - loss: 1.7715
16/16 [==============================] - 0s 792us/step - loss: 1.7674
16/16 [==============================] - 0s 792us/step - loss: 1.7659

Testing for epoch 23 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 921us/step - loss: 0.1535
16/16 [==============================] - 0s 2ms/step - loss: 1.4858
16/16 [==============================] - 0s 820us/step - loss: 1.7596
16/16 [==============================] - 0s 800us/step - loss: 1.8211
16/16 [==============================] - 0s 793us/step - loss: 1.8196
16/16 [==============================] - 0s 2ms/step - loss: 1.8077
16/16 [==============================] - 0s 840us/step - loss: 1.7924
16/16 [==============================] - 0s 870us/step - loss: 1.7827
16/16 [==============================] - 0s 1ms/step - loss: 1.7786
16/16 [==============================] - 0s 2ms/step - loss: 1.7772

Testing for epoch 23 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 849us/step - loss: 0.1485
16/16 [==============================] - 0s 2ms/step - loss: 1.4917
16/16 [==============================] - 0s 1ms/step - loss: 1.7753
16/16 [==============================] - 0s 901us/step - loss: 1.8377
16/16 [==============================] - 0s 1ms/step - loss: 1.8375
16/16 [==============================] - 0s 1ms/step - loss: 1.8268
16/16 [==============================] - 0s 2ms/step - loss: 1.8121
16/16 [==============================] - 0s 901us/step - loss: 1.8024
16/16 [==============================] - 0s 909us/step - loss: 1.7981
16/16 [==============================] - 0s 815us/step - loss: 1.7966

Testing for epoch 23 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1484
16/16 [==============================] - 0s 1ms/step - loss: 1.4544
16/16 [==============================] - 0s 846us/step - loss: 1.7244
16/16 [==============================] - 0s 2ms/step - loss: 1.7837
16/16 [==============================] - 0s 2ms/step - loss: 1.7832
16/16 [==============================] - 0s 856us/step - loss: 1.7717
16/16 [==============================] - 0s 811us/step - loss: 1.7570
16/16 [==============================] - 0s 2ms/step - loss: 1.7477
16/16 [==============================] - 0s 2ms/step - loss: 1.7438
16/16 [==============================] - 0s 2ms/step - loss: 1.7425
Epoch 24 of 60

Testing for epoch 24 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.1439
16/16 [==============================] - 0s 2ms/step - loss: 1.4811
16/16 [==============================] - 0s 876us/step - loss: 1.7631
16/16 [==============================] - 0s 793us/step - loss: 1.8232
16/16 [==============================] - 0s 2ms/step - loss: 1.8216
16/16 [==============================] - 0s 2ms/step - loss: 1.8096
16/16 [==============================] - 0s 2ms/step - loss: 1.7938
16/16 [==============================] - 0s 2ms/step - loss: 1.7839
16/16 [==============================] - 0s 2ms/step - loss: 1.7796
16/16 [==============================] - 0s 826us/step - loss: 1.7781

Testing for epoch 24 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 941us/step - loss: 0.1409
16/16 [==============================] - 0s 997us/step - loss: 1.5099
16/16 [==============================] - 0s 2ms/step - loss: 1.8010
16/16 [==============================] - 0s 2ms/step - loss: 1.8576
16/16 [==============================] - 0s 861us/step - loss: 1.8541
16/16 [==============================] - 0s 1ms/step - loss: 1.8404
16/16 [==============================] - 0s 2ms/step - loss: 1.8234
16/16 [==============================] - 0s 828us/step - loss: 1.8131
16/16 [==============================] - 0s 1ms/step - loss: 1.8088
16/16 [==============================] - 0s 850us/step - loss: 1.8075

Testing for epoch 24 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 825us/step - loss: 0.1407
16/16 [==============================] - 0s 784us/step - loss: 1.5069
16/16 [==============================] - 0s 802us/step - loss: 1.8032
16/16 [==============================] - 0s 1ms/step - loss: 1.8644
16/16 [==============================] - 0s 2ms/step - loss: 1.8632
16/16 [==============================] - 0s 835us/step - loss: 1.8512
16/16 [==============================] - 0s 836us/step - loss: 1.8357
16/16 [==============================] - 0s 2ms/step - loss: 1.8259
16/16 [==============================] - 0s 2ms/step - loss: 1.8218
16/16 [==============================] - 0s 966us/step - loss: 1.8204

Testing for epoch 24 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1394
16/16 [==============================] - 0s 844us/step - loss: 1.4794
16/16 [==============================] - 0s 804us/step - loss: 1.7777
16/16 [==============================] - 0s 740us/step - loss: 1.8444
16/16 [==============================] - 0s 774us/step - loss: 1.8460
16/16 [==============================] - 0s 799us/step - loss: 1.8363
16/16 [==============================] - 0s 795us/step - loss: 1.8218
16/16 [==============================] - 0s 774us/step - loss: 1.8121
16/16 [==============================] - 0s 805us/step - loss: 1.8078
16/16 [==============================] - 0s 791us/step - loss: 1.8063

Testing for epoch 24 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 834us/step - loss: 0.1379
16/16 [==============================] - 0s 814us/step - loss: 1.5212
16/16 [==============================] - 0s 1ms/step - loss: 1.8264
16/16 [==============================] - 0s 2ms/step - loss: 1.8886
16/16 [==============================] - 0s 801us/step - loss: 1.8851
16/16 [==============================] - 0s 798us/step - loss: 1.8698
16/16 [==============================] - 0s 1ms/step - loss: 1.8519
16/16 [==============================] - 0s 816us/step - loss: 1.8410
16/16 [==============================] - 0s 799us/step - loss: 1.8365
16/16 [==============================] - 0s 2ms/step - loss: 1.8350
Epoch 25 of 60

Testing for epoch 25 index 1:
79/79 [==============================] - 0s 2ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1387
16/16 [==============================] - 0s 944us/step - loss: 1.5261
16/16 [==============================] - 0s 789us/step - loss: 1.8311
16/16 [==============================] - 0s 2ms/step - loss: 1.8989
16/16 [==============================] - 0s 822us/step - loss: 1.8986
16/16 [==============================] - 0s 811us/step - loss: 1.8862
16/16 [==============================] - 0s 784us/step - loss: 1.8695
16/16 [==============================] - 0s 783us/step - loss: 1.8590
16/16 [==============================] - 0s 807us/step - loss: 1.8545
16/16 [==============================] - 0s 806us/step - loss: 1.8530

Testing for epoch 25 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1400
16/16 [==============================] - 0s 879us/step - loss: 1.5416
16/16 [==============================] - 0s 2ms/step - loss: 1.8371
16/16 [==============================] - 0s 877us/step - loss: 1.8998
16/16 [==============================] - 0s 2ms/step - loss: 1.8958
16/16 [==============================] - 0s 1ms/step - loss: 1.8811
16/16 [==============================] - 0s 1ms/step - loss: 1.8632
16/16 [==============================] - 0s 813us/step - loss: 1.8526
16/16 [==============================] - 0s 2ms/step - loss: 1.8483
16/16 [==============================] - 0s 2ms/step - loss: 1.8469

Testing for epoch 25 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 881us/step - loss: 0.1348
16/16 [==============================] - 0s 922us/step - loss: 1.5746
16/16 [==============================] - 0s 867us/step - loss: 1.8787
16/16 [==============================] - 0s 2ms/step - loss: 1.9429
16/16 [==============================] - 0s 1ms/step - loss: 1.9376
16/16 [==============================] - 0s 2ms/step - loss: 1.9217
16/16 [==============================] - 0s 825us/step - loss: 1.9027
16/16 [==============================] - 0s 2ms/step - loss: 1.8917
16/16 [==============================] - 0s 2ms/step - loss: 1.8873
16/16 [==============================] - 0s 970us/step - loss: 1.8860

Testing for epoch 25 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.1318
16/16 [==============================] - 0s 2ms/step - loss: 1.5680
16/16 [==============================] - 0s 937us/step - loss: 1.8780
16/16 [==============================] - 0s 2ms/step - loss: 1.9470
16/16 [==============================] - 0s 937us/step - loss: 1.9437
16/16 [==============================] - 0s 2ms/step - loss: 1.9288
16/16 [==============================] - 0s 1ms/step - loss: 1.9096
16/16 [==============================] - 0s 2ms/step - loss: 1.8985
16/16 [==============================] - 0s 1ms/step - loss: 1.8940
16/16 [==============================] - 0s 2ms/step - loss: 1.8925

Testing for epoch 25 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.1323
16/16 [==============================] - 0s 785us/step - loss: 1.5430
16/16 [==============================] - 0s 2ms/step - loss: 1.8371
16/16 [==============================] - 0s 2ms/step - loss: 1.9035
16/16 [==============================] - 0s 1ms/step - loss: 1.9006
16/16 [==============================] - 0s 891us/step - loss: 1.8875
16/16 [==============================] - 0s 1ms/step - loss: 1.8704
16/16 [==============================] - 0s 2ms/step - loss: 1.8601
16/16 [==============================] - 0s 788us/step - loss: 1.8559
16/16 [==============================] - 0s 785us/step - loss: 1.8545
Epoch 26 of 60

Testing for epoch 26 index 1:
79/79 [==============================] - 0s 548us/step
16/16 [==============================] - 0s 808us/step - loss: 0.1312
16/16 [==============================] - 0s 2ms/step - loss: 1.5662
16/16 [==============================] - 0s 1ms/step - loss: 1.8601
16/16 [==============================] - 0s 1ms/step - loss: 1.9226
16/16 [==============================] - 0s 797us/step - loss: 1.9149
16/16 [==============================] - 0s 2ms/step - loss: 1.8993
16/16 [==============================] - 0s 791us/step - loss: 1.8800
16/16 [==============================] - 0s 2ms/step - loss: 1.8689
16/16 [==============================] - 0s 779us/step - loss: 1.8644
16/16 [==============================] - 0s 2ms/step - loss: 1.8630

Testing for epoch 26 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1308
16/16 [==============================] - 0s 2ms/step - loss: 1.5812
16/16 [==============================] - 0s 1ms/step - loss: 1.8801
16/16 [==============================] - 0s 831us/step - loss: 1.9452
16/16 [==============================] - 0s 828us/step - loss: 1.9388
16/16 [==============================] - 0s 2ms/step - loss: 1.9232
16/16 [==============================] - 0s 1ms/step - loss: 1.9043
16/16 [==============================] - 0s 2ms/step - loss: 1.8934
16/16 [==============================] - 0s 2ms/step - loss: 1.8890
16/16 [==============================] - 0s 2ms/step - loss: 1.8877

Testing for epoch 26 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1296
16/16 [==============================] - 0s 920us/step - loss: 1.5545
16/16 [==============================] - 0s 2ms/step - loss: 1.8535
16/16 [==============================] - 0s 2ms/step - loss: 1.9218
16/16 [==============================] - 0s 866us/step - loss: 1.9171
16/16 [==============================] - 0s 2ms/step - loss: 1.9035
16/16 [==============================] - 0s 894us/step - loss: 1.8858
16/16 [==============================] - 0s 839us/step - loss: 1.8753
16/16 [==============================] - 0s 2ms/step - loss: 1.8710
16/16 [==============================] - 0s 1ms/step - loss: 1.8696

Testing for epoch 26 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 930us/step - loss: 0.1253
16/16 [==============================] - 0s 784us/step - loss: 1.5799
16/16 [==============================] - 0s 2ms/step - loss: 1.8860
16/16 [==============================] - 0s 859us/step - loss: 1.9532
16/16 [==============================] - 0s 797us/step - loss: 1.9468
16/16 [==============================] - 0s 792us/step - loss: 1.9325
16/16 [==============================] - 0s 2ms/step - loss: 1.9138
16/16 [==============================] - 0s 827us/step - loss: 1.9027
16/16 [==============================] - 0s 2ms/step - loss: 1.8982
16/16 [==============================] - 0s 2ms/step - loss: 1.8967

Testing for epoch 26 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.1269
16/16 [==============================] - 0s 1ms/step - loss: 1.5803
16/16 [==============================] - 0s 1ms/step - loss: 1.8920
16/16 [==============================] - 0s 970us/step - loss: 1.9613
16/16 [==============================] - 0s 2ms/step - loss: 1.9560
16/16 [==============================] - 0s 834us/step - loss: 1.9420
16/16 [==============================] - 0s 837us/step - loss: 1.9229
16/16 [==============================] - 0s 2ms/step - loss: 1.9119
16/16 [==============================] - 0s 1ms/step - loss: 1.9075
16/16 [==============================] - 0s 2ms/step - loss: 1.9061
Epoch 27 of 60

Testing for epoch 27 index 1:
79/79 [==============================] - 0s 2ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1268
16/16 [==============================] - 0s 928us/step - loss: 1.5500
16/16 [==============================] - 0s 2ms/step - loss: 1.8525
16/16 [==============================] - 0s 2ms/step - loss: 1.9169
16/16 [==============================] - 0s 852us/step - loss: 1.9095
16/16 [==============================] - 0s 2ms/step - loss: 1.8953
16/16 [==============================] - 0s 896us/step - loss: 1.8767
16/16 [==============================] - 0s 1ms/step - loss: 1.8661
16/16 [==============================] - 0s 2ms/step - loss: 1.8618
16/16 [==============================] - 0s 848us/step - loss: 1.8605

Testing for epoch 27 index 2:
79/79 [==============================] - 0s 532us/step
16/16 [==============================] - 0s 1ms/step - loss: 0.1281
16/16 [==============================] - 0s 796us/step - loss: 1.5308
16/16 [==============================] - 0s 798us/step - loss: 1.8295
16/16 [==============================] - 0s 807us/step - loss: 1.8894
16/16 [==============================] - 0s 2ms/step - loss: 1.8819
16/16 [==============================] - 0s 861us/step - loss: 1.8655
16/16 [==============================] - 0s 2ms/step - loss: 1.8453
16/16 [==============================] - 0s 1ms/step - loss: 1.8344
16/16 [==============================] - 0s 813us/step - loss: 1.8300
16/16 [==============================] - 0s 898us/step - loss: 1.8287

Testing for epoch 27 index 3:
79/79 [==============================] - 0s 2ms/step
16/16 [==============================] - 0s 945us/step - loss: 0.1295
16/16 [==============================] - 0s 857us/step - loss: 1.5644
16/16 [==============================] - 0s 803us/step - loss: 1.8682
16/16 [==============================] - 0s 830us/step - loss: 1.9264
16/16 [==============================] - 0s 2ms/step - loss: 1.9168
16/16 [==============================] - 0s 2ms/step - loss: 1.8978
16/16 [==============================] - 0s 2ms/step - loss: 1.8759
16/16 [==============================] - 0s 1ms/step - loss: 1.8643
16/16 [==============================] - 0s 1ms/step - loss: 1.8598
16/16 [==============================] - 0s 2ms/step - loss: 1.8584

Testing for epoch 27 index 4:
79/79 [==============================] - 0s 838us/step
16/16 [==============================] - 0s 893us/step - loss: 0.1229
16/16 [==============================] - 0s 843us/step - loss: 1.6017
16/16 [==============================] - 0s 2ms/step - loss: 1.9158
16/16 [==============================] - 0s 1ms/step - loss: 1.9750
16/16 [==============================] - 0s 842us/step - loss: 1.9655
16/16 [==============================] - 0s 1ms/step - loss: 1.9476
16/16 [==============================] - 0s 864us/step - loss: 1.9272
16/16 [==============================] - 0s 1ms/step - loss: 1.9159
16/16 [==============================] - 0s 2ms/step - loss: 1.9113
16/16 [==============================] - 0s 1ms/step - loss: 1.9099

Testing for epoch 27 index 5:
79/79 [==============================] - 0s 623us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1209
16/16 [==============================] - 0s 2ms/step - loss: 1.6582
16/16 [==============================] - 0s 1ms/step - loss: 1.9963
16/16 [==============================] - 0s 815us/step - loss: 2.0588
16/16 [==============================] - 0s 809us/step - loss: 2.0491
16/16 [==============================] - 0s 902us/step - loss: 2.0283
16/16 [==============================] - 0s 2ms/step - loss: 2.0048
16/16 [==============================] - 0s 785us/step - loss: 1.9922
16/16 [==============================] - 0s 2ms/step - loss: 1.9874
16/16 [==============================] - 0s 794us/step - loss: 1.9860
Epoch 28 of 60

Testing for epoch 28 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1205
16/16 [==============================] - 0s 2ms/step - loss: 1.6297
16/16 [==============================] - 0s 1ms/step - loss: 1.9572
16/16 [==============================] - 0s 2ms/step - loss: 2.0155
16/16 [==============================] - 0s 2ms/step - loss: 2.0083
16/16 [==============================] - 0s 1ms/step - loss: 1.9886
16/16 [==============================] - 0s 2ms/step - loss: 1.9666
16/16 [==============================] - 0s 845us/step - loss: 1.9547
16/16 [==============================] - 0s 813us/step - loss: 1.9500
16/16 [==============================] - 0s 819us/step - loss: 1.9486

Testing for epoch 28 index 2:
79/79 [==============================] - 0s 965us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1231
16/16 [==============================] - 0s 2ms/step - loss: 1.5918
16/16 [==============================] - 0s 2ms/step - loss: 1.9122
16/16 [==============================] - 0s 2ms/step - loss: 1.9660
16/16 [==============================] - 0s 990us/step - loss: 1.9574
16/16 [==============================] - 0s 1ms/step - loss: 1.9368
16/16 [==============================] - 0s 2ms/step - loss: 1.9143
16/16 [==============================] - 0s 2ms/step - loss: 1.9024
16/16 [==============================] - 0s 864us/step - loss: 1.8977
16/16 [==============================] - 0s 841us/step - loss: 1.8962

Testing for epoch 28 index 3:
79/79 [==============================] - 0s 952us/step
16/16 [==============================] - 0s 806us/step - loss: 0.1199
16/16 [==============================] - 0s 2ms/step - loss: 1.6289
16/16 [==============================] - 0s 807us/step - loss: 1.9569
16/16 [==============================] - 0s 791us/step - loss: 2.0083
16/16 [==============================] - 0s 808us/step - loss: 1.9981
16/16 [==============================] - 0s 2ms/step - loss: 1.9776
16/16 [==============================] - 0s 829us/step - loss: 1.9548
16/16 [==============================] - 0s 2ms/step - loss: 1.9428
16/16 [==============================] - 0s 822us/step - loss: 1.9382
16/16 [==============================] - 0s 962us/step - loss: 1.9368

Testing for epoch 28 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 885us/step - loss: 0.1181
16/16 [==============================] - 0s 2ms/step - loss: 1.6031
16/16 [==============================] - 0s 2ms/step - loss: 1.9313
16/16 [==============================] - 0s 2ms/step - loss: 1.9866
16/16 [==============================] - 0s 2ms/step - loss: 1.9794
16/16 [==============================] - 0s 925us/step - loss: 1.9606
16/16 [==============================] - 0s 807us/step - loss: 1.9384
16/16 [==============================] - 0s 920us/step - loss: 1.9267
16/16 [==============================] - 0s 2ms/step - loss: 1.9220
16/16 [==============================] - 0s 800us/step - loss: 1.9206

Testing for epoch 28 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 814us/step - loss: 0.1194
16/16 [==============================] - 0s 2ms/step - loss: 1.6061
16/16 [==============================] - 0s 2ms/step - loss: 1.9314
16/16 [==============================] - 0s 2ms/step - loss: 1.9818
16/16 [==============================] - 0s 2ms/step - loss: 1.9704
16/16 [==============================] - 0s 872us/step - loss: 1.9494
16/16 [==============================] - 0s 2ms/step - loss: 1.9268
16/16 [==============================] - 0s 857us/step - loss: 1.9151
16/16 [==============================] - 0s 2ms/step - loss: 1.9106
16/16 [==============================] - 0s 873us/step - loss: 1.9093
Epoch 29 of 60

Testing for epoch 29 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1170
16/16 [==============================] - 0s 1ms/step - loss: 1.5952
16/16 [==============================] - 0s 2ms/step - loss: 1.9156
16/16 [==============================] - 0s 2ms/step - loss: 1.9670
16/16 [==============================] - 0s 1ms/step - loss: 1.9563
16/16 [==============================] - 0s 962us/step - loss: 1.9371
16/16 [==============================] - 0s 2ms/step - loss: 1.9153
16/16 [==============================] - 0s 849us/step - loss: 1.9039
16/16 [==============================] - 0s 1ms/step - loss: 1.8994
16/16 [==============================] - 0s 2ms/step - loss: 1.8980

Testing for epoch 29 index 2:
79/79 [==============================] - 0s 2ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1179
16/16 [==============================] - 0s 2ms/step - loss: 1.6545
16/16 [==============================] - 0s 2ms/step - loss: 1.9932
16/16 [==============================] - 0s 2ms/step - loss: 2.0482
16/16 [==============================] - 0s 920us/step - loss: 2.0365
16/16 [==============================] - 0s 2ms/step - loss: 2.0159
16/16 [==============================] - 0s 1ms/step - loss: 1.9922
16/16 [==============================] - 0s 816us/step - loss: 1.9800
16/16 [==============================] - 0s 804us/step - loss: 1.9752
16/16 [==============================] - 0s 911us/step - loss: 1.9737

Testing for epoch 29 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.1166
16/16 [==============================] - 0s 820us/step - loss: 1.6659
16/16 [==============================] - 0s 2ms/step - loss: 2.0026
16/16 [==============================] - 0s 2ms/step - loss: 2.0614
16/16 [==============================] - 0s 2ms/step - loss: 2.0491
16/16 [==============================] - 0s 897us/step - loss: 2.0274
16/16 [==============================] - 0s 1ms/step - loss: 2.0028
16/16 [==============================] - 0s 2ms/step - loss: 1.9905
16/16 [==============================] - 0s 1ms/step - loss: 1.9859
16/16 [==============================] - 0s 867us/step - loss: 1.9846

Testing for epoch 29 index 4:
79/79 [==============================] - 0s 544us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1145
16/16 [==============================] - 0s 803us/step - loss: 1.6746
16/16 [==============================] - 0s 953us/step - loss: 2.0219
16/16 [==============================] - 0s 2ms/step - loss: 2.0813
16/16 [==============================] - 0s 1ms/step - loss: 2.0677
16/16 [==============================] - 0s 2ms/step - loss: 2.0448
16/16 [==============================] - 0s 2ms/step - loss: 2.0186
16/16 [==============================] - 0s 2ms/step - loss: 2.0056
16/16 [==============================] - 0s 791us/step - loss: 2.0007
16/16 [==============================] - 0s 1ms/step - loss: 1.9992

Testing for epoch 29 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 958us/step - loss: 0.1158
16/16 [==============================] - 0s 797us/step - loss: 1.6228
16/16 [==============================] - 0s 912us/step - loss: 1.9529
16/16 [==============================] - 0s 822us/step - loss: 2.0115
16/16 [==============================] - 0s 2ms/step - loss: 1.9991
16/16 [==============================] - 0s 834us/step - loss: 1.9784
16/16 [==============================] - 0s 2ms/step - loss: 1.9536
16/16 [==============================] - 0s 2ms/step - loss: 1.9414
16/16 [==============================] - 0s 858us/step - loss: 1.9367
16/16 [==============================] - 0s 783us/step - loss: 1.9354
Epoch 30 of 60

Testing for epoch 30 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1159
16/16 [==============================] - 0s 2ms/step - loss: 1.6216
16/16 [==============================] - 0s 1ms/step - loss: 1.9505
16/16 [==============================] - 0s 2ms/step - loss: 2.0062
16/16 [==============================] - 0s 885us/step - loss: 1.9926
16/16 [==============================] - 0s 998us/step - loss: 1.9723
16/16 [==============================] - 0s 850us/step - loss: 1.9478
16/16 [==============================] - 0s 772us/step - loss: 1.9356
16/16 [==============================] - 0s 786us/step - loss: 1.9310
16/16 [==============================] - 0s 787us/step - loss: 1.9297

Testing for epoch 30 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 898us/step - loss: 0.1135
16/16 [==============================] - 0s 1ms/step - loss: 1.6538
16/16 [==============================] - 0s 849us/step - loss: 1.9883
16/16 [==============================] - 0s 794us/step - loss: 2.0417
16/16 [==============================] - 0s 797us/step - loss: 2.0248
16/16 [==============================] - 0s 811us/step - loss: 2.0018
16/16 [==============================] - 0s 803us/step - loss: 1.9755
16/16 [==============================] - 0s 788us/step - loss: 1.9627
16/16 [==============================] - 0s 847us/step - loss: 1.9580
16/16 [==============================] - 0s 858us/step - loss: 1.9566

Testing for epoch 30 index 3:
79/79 [==============================] - 0s 904us/step
16/16 [==============================] - 0s 819us/step - loss: 0.1135
16/16 [==============================] - 0s 814us/step - loss: 1.6788
16/16 [==============================] - 0s 794us/step - loss: 2.0212
16/16 [==============================] - 0s 784us/step - loss: 2.0796
16/16 [==============================] - 0s 794us/step - loss: 2.0633
16/16 [==============================] - 0s 778us/step - loss: 2.0402
16/16 [==============================] - 0s 778us/step - loss: 2.0132
16/16 [==============================] - 0s 778us/step - loss: 2.0001
16/16 [==============================] - 0s 776us/step - loss: 1.9951
16/16 [==============================] - 0s 796us/step - loss: 1.9937

Testing for epoch 30 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 810us/step - loss: 0.1126
16/16 [==============================] - 0s 800us/step - loss: 1.6895
16/16 [==============================] - 0s 808us/step - loss: 2.0271
16/16 [==============================] - 0s 997us/step - loss: 2.0808
16/16 [==============================] - 0s 785us/step - loss: 2.0617
16/16 [==============================] - 0s 810us/step - loss: 2.0363
16/16 [==============================] - 0s 1ms/step - loss: 2.0086
16/16 [==============================] - 0s 796us/step - loss: 1.9954
16/16 [==============================] - 0s 2ms/step - loss: 1.9906
16/16 [==============================] - 0s 1ms/step - loss: 1.9892

Testing for epoch 30 index 5:
79/79 [==============================] - 0s 537us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1123
16/16 [==============================] - 0s 814us/step - loss: 1.6778
16/16 [==============================] - 0s 2ms/step - loss: 2.0144
16/16 [==============================] - 0s 2ms/step - loss: 2.0707
16/16 [==============================] - 0s 1ms/step - loss: 2.0538
16/16 [==============================] - 0s 1ms/step - loss: 2.0315
16/16 [==============================] - 0s 854us/step - loss: 2.0061
16/16 [==============================] - 0s 818us/step - loss: 1.9937
16/16 [==============================] - 0s 2ms/step - loss: 1.9890
16/16 [==============================] - 0s 2ms/step - loss: 1.9876
Epoch 31 of 60

Testing for epoch 31 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 890us/step - loss: 0.1101
16/16 [==============================] - 0s 808us/step - loss: 1.6815
16/16 [==============================] - 0s 792us/step - loss: 2.0160
16/16 [==============================] - 0s 811us/step - loss: 2.0706
16/16 [==============================] - 0s 798us/step - loss: 2.0509
16/16 [==============================] - 0s 2ms/step - loss: 2.0265
16/16 [==============================] - 0s 2ms/step - loss: 2.0000
16/16 [==============================] - 0s 845us/step - loss: 1.9874
16/16 [==============================] - 0s 796us/step - loss: 1.9828
16/16 [==============================] - 0s 2ms/step - loss: 1.9814

Testing for epoch 31 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 973us/step - loss: 0.1125
16/16 [==============================] - 0s 849us/step - loss: 1.6742
16/16 [==============================] - 0s 1ms/step - loss: 2.0050
16/16 [==============================] - 0s 895us/step - loss: 2.0596
16/16 [==============================] - 0s 795us/step - loss: 2.0400
16/16 [==============================] - 0s 1ms/step - loss: 2.0148
16/16 [==============================] - 0s 2ms/step - loss: 1.9884
16/16 [==============================] - 0s 2ms/step - loss: 1.9757
16/16 [==============================] - 0s 802us/step - loss: 1.9711
16/16 [==============================] - 0s 796us/step - loss: 1.9697

Testing for epoch 31 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 932us/step - loss: 0.1094
16/16 [==============================] - 0s 1ms/step - loss: 1.6789
16/16 [==============================] - 0s 2ms/step - loss: 2.0053
16/16 [==============================] - 0s 824us/step - loss: 2.0575
16/16 [==============================] - 0s 787us/step - loss: 2.0375
16/16 [==============================] - 0s 824us/step - loss: 2.0138
16/16 [==============================] - 0s 2ms/step - loss: 1.9874
16/16 [==============================] - 0s 2ms/step - loss: 1.9748
16/16 [==============================] - 0s 1ms/step - loss: 1.9702
16/16 [==============================] - 0s 858us/step - loss: 1.9689

Testing for epoch 31 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1056
16/16 [==============================] - 0s 2ms/step - loss: 1.6807
16/16 [==============================] - 0s 2ms/step - loss: 2.0141
16/16 [==============================] - 0s 2ms/step - loss: 2.0715
16/16 [==============================] - 0s 2ms/step - loss: 2.0530
16/16 [==============================] - 0s 2ms/step - loss: 2.0299
16/16 [==============================] - 0s 2ms/step - loss: 2.0042
16/16 [==============================] - 0s 2ms/step - loss: 1.9918
16/16 [==============================] - 0s 1ms/step - loss: 1.9873
16/16 [==============================] - 0s 891us/step - loss: 1.9860

Testing for epoch 31 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 843us/step - loss: 0.1083
16/16 [==============================] - 0s 1ms/step - loss: 1.6546
16/16 [==============================] - 0s 804us/step - loss: 1.9743
16/16 [==============================] - 0s 2ms/step - loss: 2.0268
16/16 [==============================] - 0s 2ms/step - loss: 2.0072
16/16 [==============================] - 0s 812us/step - loss: 1.9828
16/16 [==============================] - 0s 803us/step - loss: 1.9560
16/16 [==============================] - 0s 1ms/step - loss: 1.9432
16/16 [==============================] - 0s 813us/step - loss: 1.9386
16/16 [==============================] - 0s 1ms/step - loss: 1.9372
Epoch 32 of 60

Testing for epoch 32 index 1:
79/79 [==============================] - 0s 522us/step
16/16 [==============================] - 0s 823us/step - loss: 0.1059
16/16 [==============================] - 0s 825us/step - loss: 1.7084
16/16 [==============================] - 0s 802us/step - loss: 2.0421
16/16 [==============================] - 0s 2ms/step - loss: 2.0994
16/16 [==============================] - 0s 848us/step - loss: 2.0779
16/16 [==============================] - 0s 798us/step - loss: 2.0520
16/16 [==============================] - 0s 767us/step - loss: 2.0243
16/16 [==============================] - 0s 789us/step - loss: 2.0112
16/16 [==============================] - 0s 804us/step - loss: 2.0065
16/16 [==============================] - 0s 806us/step - loss: 2.0051

Testing for epoch 32 index 2:
79/79 [==============================] - 0s 806us/step
16/16 [==============================] - 0s 789us/step - loss: 0.1044
16/16 [==============================] - 0s 789us/step - loss: 1.7390
16/16 [==============================] - 0s 790us/step - loss: 2.0806
16/16 [==============================] - 0s 801us/step - loss: 2.1353
16/16 [==============================] - 0s 1ms/step - loss: 2.1106
16/16 [==============================] - 0s 2ms/step - loss: 2.0819
16/16 [==============================] - 0s 2ms/step - loss: 2.0526
16/16 [==============================] - 0s 1ms/step - loss: 2.0390
16/16 [==============================] - 0s 2ms/step - loss: 2.0340
16/16 [==============================] - 0s 2ms/step - loss: 2.0326

Testing for epoch 32 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1093
16/16 [==============================] - 0s 2ms/step - loss: 1.6886
16/16 [==============================] - 0s 2ms/step - loss: 2.0213
16/16 [==============================] - 0s 2ms/step - loss: 2.0751
16/16 [==============================] - 0s 2ms/step - loss: 2.0513
16/16 [==============================] - 0s 2ms/step - loss: 2.0243
16/16 [==============================] - 0s 1ms/step - loss: 1.9958
16/16 [==============================] - 0s 1ms/step - loss: 1.9825
16/16 [==============================] - 0s 842us/step - loss: 1.9776
16/16 [==============================] - 0s 803us/step - loss: 1.9762

Testing for epoch 32 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1047
16/16 [==============================] - 0s 834us/step - loss: 1.7320
16/16 [==============================] - 0s 870us/step - loss: 2.0796
16/16 [==============================] - 0s 2ms/step - loss: 2.1341
16/16 [==============================] - 0s 2ms/step - loss: 2.1104
16/16 [==============================] - 0s 820us/step - loss: 2.0836
16/16 [==============================] - 0s 2ms/step - loss: 2.0549
16/16 [==============================] - 0s 951us/step - loss: 2.0415
16/16 [==============================] - 0s 812us/step - loss: 2.0367
16/16 [==============================] - 0s 2ms/step - loss: 2.0354

Testing for epoch 32 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1053
16/16 [==============================] - 0s 2ms/step - loss: 1.7510
16/16 [==============================] - 0s 2ms/step - loss: 2.1031
16/16 [==============================] - 0s 897us/step - loss: 2.1564
16/16 [==============================] - 0s 901us/step - loss: 2.1313
16/16 [==============================] - 0s 2ms/step - loss: 2.1030
16/16 [==============================] - 0s 2ms/step - loss: 2.0735
16/16 [==============================] - 0s 1ms/step - loss: 2.0597
16/16 [==============================] - 0s 799us/step - loss: 2.0548
16/16 [==============================] - 0s 2ms/step - loss: 2.0533
Epoch 33 of 60

Testing for epoch 33 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1042
16/16 [==============================] - 0s 2ms/step - loss: 1.7352
16/16 [==============================] - 0s 925us/step - loss: 2.0848
16/16 [==============================] - 0s 2ms/step - loss: 2.1397
16/16 [==============================] - 0s 2ms/step - loss: 2.1162
16/16 [==============================] - 0s 2ms/step - loss: 2.0905
16/16 [==============================] - 0s 2ms/step - loss: 2.0623
16/16 [==============================] - 0s 2ms/step - loss: 2.0489
16/16 [==============================] - 0s 1ms/step - loss: 2.0440
16/16 [==============================] - 0s 910us/step - loss: 2.0426

Testing for epoch 33 index 2:
79/79 [==============================] - 0s 868us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1036
16/16 [==============================] - 0s 2ms/step - loss: 1.6918
16/16 [==============================] - 0s 837us/step - loss: 2.0308
16/16 [==============================] - 0s 2ms/step - loss: 2.0838
16/16 [==============================] - 0s 2ms/step - loss: 2.0615
16/16 [==============================] - 0s 837us/step - loss: 2.0372
16/16 [==============================] - 0s 915us/step - loss: 2.0101
16/16 [==============================] - 0s 2ms/step - loss: 1.9973
16/16 [==============================] - 0s 2ms/step - loss: 1.9927
16/16 [==============================] - 0s 2ms/step - loss: 1.9914

Testing for epoch 33 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1041
16/16 [==============================] - 0s 791us/step - loss: 1.7168
16/16 [==============================] - 0s 2ms/step - loss: 2.0616
16/16 [==============================] - 0s 805us/step - loss: 2.1092
16/16 [==============================] - 0s 799us/step - loss: 2.0820
16/16 [==============================] - 0s 812us/step - loss: 2.0542
16/16 [==============================] - 0s 799us/step - loss: 2.0252
16/16 [==============================] - 0s 784us/step - loss: 2.0118
16/16 [==============================] - 0s 791us/step - loss: 2.0070
16/16 [==============================] - 0s 781us/step - loss: 2.0057

Testing for epoch 33 index 4:
79/79 [==============================] - 0s 547us/step
16/16 [==============================] - 0s 822us/step - loss: 0.1023
16/16 [==============================] - 0s 2ms/step - loss: 1.7233
16/16 [==============================] - 0s 808us/step - loss: 2.0734
16/16 [==============================] - 0s 796us/step - loss: 2.1271
16/16 [==============================] - 0s 803us/step - loss: 2.1028
16/16 [==============================] - 0s 796us/step - loss: 2.0779
16/16 [==============================] - 0s 2ms/step - loss: 2.0499
16/16 [==============================] - 0s 783us/step - loss: 2.0367
16/16 [==============================] - 0s 829us/step - loss: 2.0319
16/16 [==============================] - 0s 875us/step - loss: 2.0305

Testing for epoch 33 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1024
16/16 [==============================] - 0s 2ms/step - loss: 1.7279
16/16 [==============================] - 0s 943us/step - loss: 2.0767
16/16 [==============================] - 0s 837us/step - loss: 2.1247
16/16 [==============================] - 0s 2ms/step - loss: 2.0985
16/16 [==============================] - 0s 2ms/step - loss: 2.0716
16/16 [==============================] - 0s 2ms/step - loss: 2.0428
16/16 [==============================] - 0s 934us/step - loss: 2.0295
16/16 [==============================] - 0s 2ms/step - loss: 2.0247
16/16 [==============================] - 0s 911us/step - loss: 2.0233
Epoch 34 of 60

Testing for epoch 34 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.1033
16/16 [==============================] - 0s 2ms/step - loss: 1.6971
16/16 [==============================] - 0s 2ms/step - loss: 2.0431
16/16 [==============================] - 0s 2ms/step - loss: 2.0937
16/16 [==============================] - 0s 1ms/step - loss: 2.0696
16/16 [==============================] - 0s 1ms/step - loss: 2.0442
16/16 [==============================] - 0s 811us/step - loss: 2.0163
16/16 [==============================] - 0s 769us/step - loss: 2.0032
16/16 [==============================] - 0s 2ms/step - loss: 1.9984
16/16 [==============================] - 0s 830us/step - loss: 1.9969

Testing for epoch 34 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 809us/step - loss: 0.1011
16/16 [==============================] - 0s 2ms/step - loss: 1.7431
16/16 [==============================] - 0s 1ms/step - loss: 2.0892
16/16 [==============================] - 0s 2ms/step - loss: 2.1349
16/16 [==============================] - 0s 878us/step - loss: 2.1070
16/16 [==============================] - 0s 2ms/step - loss: 2.0785
16/16 [==============================] - 0s 1ms/step - loss: 2.0484
16/16 [==============================] - 0s 2ms/step - loss: 2.0346
16/16 [==============================] - 0s 2ms/step - loss: 2.0298
16/16 [==============================] - 0s 882us/step - loss: 2.0284

Testing for epoch 34 index 3:
79/79 [==============================] - 0s 527us/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0980
16/16 [==============================] - 0s 1ms/step - loss: 1.7630
16/16 [==============================] - 0s 2ms/step - loss: 2.1150
16/16 [==============================] - 0s 2ms/step - loss: 2.1627
16/16 [==============================] - 0s 2ms/step - loss: 2.1348
16/16 [==============================] - 0s 1ms/step - loss: 2.1076
16/16 [==============================] - 0s 944us/step - loss: 2.0785
16/16 [==============================] - 0s 1ms/step - loss: 2.0650
16/16 [==============================] - 0s 1ms/step - loss: 2.0601
16/16 [==============================] - 0s 837us/step - loss: 2.0587

Testing for epoch 34 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 805us/step - loss: 0.1000
16/16 [==============================] - 0s 1ms/step - loss: 1.7837
16/16 [==============================] - 0s 836us/step - loss: 2.1409
16/16 [==============================] - 0s 768us/step - loss: 2.1874
16/16 [==============================] - 0s 795us/step - loss: 2.1579
16/16 [==============================] - 0s 838us/step - loss: 2.1286
16/16 [==============================] - 0s 1ms/step - loss: 2.0982
16/16 [==============================] - 0s 1ms/step - loss: 2.0842
16/16 [==============================] - 0s 2ms/step - loss: 2.0792
16/16 [==============================] - 0s 2ms/step - loss: 2.0778

Testing for epoch 34 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0980
16/16 [==============================] - 0s 998us/step - loss: 1.7543
16/16 [==============================] - 0s 847us/step - loss: 2.1009
16/16 [==============================] - 0s 919us/step - loss: 2.1467
16/16 [==============================] - 0s 2ms/step - loss: 2.1179
16/16 [==============================] - 0s 818us/step - loss: 2.0890
16/16 [==============================] - 0s 2ms/step - loss: 2.0592
16/16 [==============================] - 0s 2ms/step - loss: 2.0457
16/16 [==============================] - 0s 827us/step - loss: 2.0409
16/16 [==============================] - 0s 1ms/step - loss: 2.0396
Epoch 35 of 60

Testing for epoch 35 index 1:
79/79 [==============================] - 0s 967us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0996
16/16 [==============================] - 0s 2ms/step - loss: 1.7749
16/16 [==============================] - 0s 1ms/step - loss: 2.1228
16/16 [==============================] - 0s 835us/step - loss: 2.1664
16/16 [==============================] - 0s 2ms/step - loss: 2.1358
16/16 [==============================] - 0s 942us/step - loss: 2.1046
16/16 [==============================] - 0s 1ms/step - loss: 2.0731
16/16 [==============================] - 0s 2ms/step - loss: 2.0590
16/16 [==============================] - 0s 2ms/step - loss: 2.0540
16/16 [==============================] - 0s 878us/step - loss: 2.0527

Testing for epoch 35 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0994
16/16 [==============================] - 0s 2ms/step - loss: 1.7644
16/16 [==============================] - 0s 788us/step - loss: 2.1129
16/16 [==============================] - 0s 800us/step - loss: 2.1556
16/16 [==============================] - 0s 2ms/step - loss: 2.1258
16/16 [==============================] - 0s 991us/step - loss: 2.0959
16/16 [==============================] - 0s 802us/step - loss: 2.0654
16/16 [==============================] - 0s 809us/step - loss: 2.0517
16/16 [==============================] - 0s 964us/step - loss: 2.0469
16/16 [==============================] - 0s 821us/step - loss: 2.0457

Testing for epoch 35 index 3:
79/79 [==============================] - 0s 957us/step
16/16 [==============================] - 0s 809us/step - loss: 0.0977
16/16 [==============================] - 0s 2ms/step - loss: 1.7360
16/16 [==============================] - 0s 776us/step - loss: 2.0806
16/16 [==============================] - 0s 821us/step - loss: 2.1208
16/16 [==============================] - 0s 797us/step - loss: 2.0925
16/16 [==============================] - 0s 806us/step - loss: 2.0644
16/16 [==============================] - 0s 835us/step - loss: 2.0354
16/16 [==============================] - 0s 794us/step - loss: 2.0221
16/16 [==============================] - 0s 2ms/step - loss: 2.0172
16/16 [==============================] - 0s 2ms/step - loss: 2.0158

Testing for epoch 35 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0945
16/16 [==============================] - 0s 825us/step - loss: 1.7867
16/16 [==============================] - 0s 2ms/step - loss: 2.1448
16/16 [==============================] - 0s 2ms/step - loss: 2.1815
16/16 [==============================] - 0s 2ms/step - loss: 2.1496
16/16 [==============================] - 0s 946us/step - loss: 2.1162
16/16 [==============================] - 0s 2ms/step - loss: 2.0834
16/16 [==============================] - 0s 862us/step - loss: 2.0689
16/16 [==============================] - 0s 881us/step - loss: 2.0639
16/16 [==============================] - 0s 817us/step - loss: 2.0626

Testing for epoch 35 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 807us/step - loss: 0.0977
16/16 [==============================] - 0s 2ms/step - loss: 1.7696
16/16 [==============================] - 0s 856us/step - loss: 2.1303
16/16 [==============================] - 0s 814us/step - loss: 2.1687
16/16 [==============================] - 0s 821us/step - loss: 2.1397
16/16 [==============================] - 0s 803us/step - loss: 2.1090
16/16 [==============================] - 0s 795us/step - loss: 2.0780
16/16 [==============================] - 0s 886us/step - loss: 2.0639
16/16 [==============================] - 0s 2ms/step - loss: 2.0589
16/16 [==============================] - 0s 801us/step - loss: 2.0574
Epoch 36 of 60

Testing for epoch 36 index 1:
79/79 [==============================] - 0s 541us/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0936
16/16 [==============================] - 0s 2ms/step - loss: 1.8110
16/16 [==============================] - 0s 2ms/step - loss: 2.1813
16/16 [==============================] - 0s 1ms/step - loss: 2.2172
16/16 [==============================] - 0s 2ms/step - loss: 2.1861
16/16 [==============================] - 0s 2ms/step - loss: 2.1521
16/16 [==============================] - 0s 1ms/step - loss: 2.1186
16/16 [==============================] - 0s 1ms/step - loss: 2.1038
16/16 [==============================] - 0s 1ms/step - loss: 2.0988
16/16 [==============================] - 0s 849us/step - loss: 2.0975

Testing for epoch 36 index 2:
79/79 [==============================] - 0s 906us/step
16/16 [==============================] - 0s 752us/step - loss: 0.0995
16/16 [==============================] - 0s 1ms/step - loss: 1.7250
16/16 [==============================] - 0s 2ms/step - loss: 2.0750
16/16 [==============================] - 0s 796us/step - loss: 2.1095
16/16 [==============================] - 0s 2ms/step - loss: 2.0832
16/16 [==============================] - 0s 2ms/step - loss: 2.0545
16/16 [==============================] - 0s 820us/step - loss: 2.0250
16/16 [==============================] - 0s 796us/step - loss: 2.0117
16/16 [==============================] - 0s 755us/step - loss: 2.0071
16/16 [==============================] - 0s 831us/step - loss: 2.0058

Testing for epoch 36 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 907us/step - loss: 0.0940
16/16 [==============================] - 0s 843us/step - loss: 1.7471
16/16 [==============================] - 0s 799us/step - loss: 2.1014
16/16 [==============================] - 0s 799us/step - loss: 2.1316
16/16 [==============================] - 0s 2ms/step - loss: 2.1030
16/16 [==============================] - 0s 785us/step - loss: 2.0721
16/16 [==============================] - 0s 2ms/step - loss: 2.0418
16/16 [==============================] - 0s 2ms/step - loss: 2.0283
16/16 [==============================] - 0s 1ms/step - loss: 2.0235
16/16 [==============================] - 0s 887us/step - loss: 2.0221

Testing for epoch 36 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0966
16/16 [==============================] - 0s 2ms/step - loss: 1.7748
16/16 [==============================] - 0s 937us/step - loss: 2.1434
16/16 [==============================] - 0s 1ms/step - loss: 2.1759
16/16 [==============================] - 0s 2ms/step - loss: 2.1474
16/16 [==============================] - 0s 2ms/step - loss: 2.1153
16/16 [==============================] - 0s 848us/step - loss: 2.0834
16/16 [==============================] - 0s 2ms/step - loss: 2.0694
16/16 [==============================] - 0s 837us/step - loss: 2.0647
16/16 [==============================] - 0s 1ms/step - loss: 2.0634

Testing for epoch 36 index 5:
79/79 [==============================] - 0s 537us/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0947
16/16 [==============================] - 0s 2ms/step - loss: 1.8190
16/16 [==============================] - 0s 2ms/step - loss: 2.1952
16/16 [==============================] - 0s 2ms/step - loss: 2.2227
16/16 [==============================] - 0s 861us/step - loss: 2.1902
16/16 [==============================] - 0s 2ms/step - loss: 2.1543
16/16 [==============================] - 0s 862us/step - loss: 2.1194
16/16 [==============================] - 0s 2ms/step - loss: 2.1042
16/16 [==============================] - 0s 2ms/step - loss: 2.0992
16/16 [==============================] - 0s 837us/step - loss: 2.0979
Epoch 37 of 60

Testing for epoch 37 index 1:
79/79 [==============================] - 0s 527us/step
16/16 [==============================] - 0s 944us/step - loss: 0.0941
16/16 [==============================] - 0s 771us/step - loss: 1.8107
16/16 [==============================] - 0s 801us/step - loss: 2.1883
16/16 [==============================] - 0s 774us/step - loss: 2.2240
16/16 [==============================] - 0s 812us/step - loss: 2.1955
16/16 [==============================] - 0s 2ms/step - loss: 2.1628
16/16 [==============================] - 0s 1ms/step - loss: 2.1296
16/16 [==============================] - 0s 1ms/step - loss: 2.1149
16/16 [==============================] - 0s 828us/step - loss: 2.1099
16/16 [==============================] - 0s 1ms/step - loss: 2.1086

Testing for epoch 37 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0915
16/16 [==============================] - 0s 1ms/step - loss: 1.8503
16/16 [==============================] - 0s 1ms/step - loss: 2.2238
16/16 [==============================] - 0s 893us/step - loss: 2.2526
16/16 [==============================] - 0s 1ms/step - loss: 2.2199
16/16 [==============================] - 0s 2ms/step - loss: 2.1840
16/16 [==============================] - 0s 882us/step - loss: 2.1493
16/16 [==============================] - 0s 870us/step - loss: 2.1342
16/16 [==============================] - 0s 2ms/step - loss: 2.1290
16/16 [==============================] - 0s 2ms/step - loss: 2.1276

Testing for epoch 37 index 3:
79/79 [==============================] - 0s 887us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0927
16/16 [==============================] - 0s 842us/step - loss: 1.8043
16/16 [==============================] - 0s 1ms/step - loss: 2.1724
16/16 [==============================] - 0s 2ms/step - loss: 2.2066
16/16 [==============================] - 0s 1ms/step - loss: 2.1779
16/16 [==============================] - 0s 903us/step - loss: 2.1450
16/16 [==============================] - 0s 1ms/step - loss: 2.1122
16/16 [==============================] - 0s 2ms/step - loss: 2.0977
16/16 [==============================] - 0s 2ms/step - loss: 2.0926
16/16 [==============================] - 0s 2ms/step - loss: 2.0912

Testing for epoch 37 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0909
16/16 [==============================] - 0s 2ms/step - loss: 1.8511
16/16 [==============================] - 0s 2ms/step - loss: 2.2282
16/16 [==============================] - 0s 2ms/step - loss: 2.2602
16/16 [==============================] - 0s 916us/step - loss: 2.2292
16/16 [==============================] - 0s 1ms/step - loss: 2.1954
16/16 [==============================] - 0s 2ms/step - loss: 2.1615
16/16 [==============================] - 0s 2ms/step - loss: 2.1467
16/16 [==============================] - 0s 934us/step - loss: 2.1417
16/16 [==============================] - 0s 849us/step - loss: 2.1403

Testing for epoch 37 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 812us/step - loss: 0.0955
16/16 [==============================] - 0s 802us/step - loss: 1.7805
16/16 [==============================] - 0s 810us/step - loss: 2.1382
16/16 [==============================] - 0s 809us/step - loss: 2.1658
16/16 [==============================] - 0s 791us/step - loss: 2.1340
16/16 [==============================] - 0s 785us/step - loss: 2.1008
16/16 [==============================] - 0s 2ms/step - loss: 2.0679
16/16 [==============================] - 0s 802us/step - loss: 2.0538
16/16 [==============================] - 0s 887us/step - loss: 2.0490
16/16 [==============================] - 0s 784us/step - loss: 2.0478
Epoch 38 of 60

Testing for epoch 38 index 1:
79/79 [==============================] - 0s 799us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0896
16/16 [==============================] - 0s 1ms/step - loss: 1.8083
16/16 [==============================] - 0s 2ms/step - loss: 2.1768
16/16 [==============================] - 0s 1ms/step - loss: 2.2060
16/16 [==============================] - 0s 2ms/step - loss: 2.1741
16/16 [==============================] - 0s 859us/step - loss: 2.1388
16/16 [==============================] - 0s 2ms/step - loss: 2.1050
16/16 [==============================] - 0s 2ms/step - loss: 2.0904
16/16 [==============================] - 0s 2ms/step - loss: 2.0854
16/16 [==============================] - 0s 2ms/step - loss: 2.0840

Testing for epoch 38 index 2:
79/79 [==============================] - 0s 608us/step
16/16 [==============================] - 0s 808us/step - loss: 0.0932
16/16 [==============================] - 0s 2ms/step - loss: 1.8485
16/16 [==============================] - 0s 801us/step - loss: 2.2167
16/16 [==============================] - 0s 919us/step - loss: 2.2375
16/16 [==============================] - 0s 2ms/step - loss: 2.2004
16/16 [==============================] - 0s 2ms/step - loss: 2.1605
16/16 [==============================] - 0s 919us/step - loss: 2.1243
16/16 [==============================] - 0s 890us/step - loss: 2.1090
16/16 [==============================] - 0s 1ms/step - loss: 2.1040
16/16 [==============================] - 0s 2ms/step - loss: 2.1027

Testing for epoch 38 index 3:
79/79 [==============================] - 0s 888us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0916
16/16 [==============================] - 0s 2ms/step - loss: 1.8503
16/16 [==============================] - 0s 848us/step - loss: 2.2202
16/16 [==============================] - 0s 2ms/step - loss: 2.2484
16/16 [==============================] - 0s 794us/step - loss: 2.2153
16/16 [==============================] - 0s 801us/step - loss: 2.1797
16/16 [==============================] - 0s 2ms/step - loss: 2.1451
16/16 [==============================] - 0s 2ms/step - loss: 2.1302
16/16 [==============================] - 0s 1ms/step - loss: 2.1251
16/16 [==============================] - 0s 1ms/step - loss: 2.1237

Testing for epoch 38 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0914
16/16 [==============================] - 0s 1ms/step - loss: 1.8048
16/16 [==============================] - 0s 2ms/step - loss: 2.1693
16/16 [==============================] - 0s 821us/step - loss: 2.2033
16/16 [==============================] - 0s 842us/step - loss: 2.1747
16/16 [==============================] - 0s 1ms/step - loss: 2.1433
16/16 [==============================] - 0s 2ms/step - loss: 2.1114
16/16 [==============================] - 0s 884us/step - loss: 2.0974
16/16 [==============================] - 0s 815us/step - loss: 2.0926
16/16 [==============================] - 0s 788us/step - loss: 2.0913

Testing for epoch 38 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0898
16/16 [==============================] - 0s 821us/step - loss: 1.8266
16/16 [==============================] - 0s 768us/step - loss: 2.1814
16/16 [==============================] - 0s 2ms/step - loss: 2.2034
16/16 [==============================] - 0s 2ms/step - loss: 2.1676
16/16 [==============================] - 0s 1ms/step - loss: 2.1288
16/16 [==============================] - 0s 2ms/step - loss: 2.0926
16/16 [==============================] - 0s 984us/step - loss: 2.0771
16/16 [==============================] - 0s 796us/step - loss: 2.0719
16/16 [==============================] - 0s 2ms/step - loss: 2.0705
Epoch 39 of 60

Testing for epoch 39 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 847us/step - loss: 0.0904
16/16 [==============================] - 0s 2ms/step - loss: 1.8701
16/16 [==============================] - 0s 2ms/step - loss: 2.2383
16/16 [==============================] - 0s 2ms/step - loss: 2.2642
16/16 [==============================] - 0s 778us/step - loss: 2.2303
16/16 [==============================] - 0s 2ms/step - loss: 2.1939
16/16 [==============================] - 0s 907us/step - loss: 2.1592
16/16 [==============================] - 0s 2ms/step - loss: 2.1443
16/16 [==============================] - 0s 1ms/step - loss: 2.1393
16/16 [==============================] - 0s 805us/step - loss: 2.1380

Testing for epoch 39 index 2:
79/79 [==============================] - 0s 525us/step
16/16 [==============================] - 0s 797us/step - loss: 0.0897
16/16 [==============================] - 0s 806us/step - loss: 1.8896
16/16 [==============================] - 0s 805us/step - loss: 2.2555
16/16 [==============================] - 0s 787us/step - loss: 2.2798
16/16 [==============================] - 0s 778us/step - loss: 2.2434
16/16 [==============================] - 0s 808us/step - loss: 2.2046
16/16 [==============================] - 0s 789us/step - loss: 2.1682
16/16 [==============================] - 0s 829us/step - loss: 2.1528
16/16 [==============================] - 0s 1ms/step - loss: 2.1478
16/16 [==============================] - 0s 815us/step - loss: 2.1466

Testing for epoch 39 index 3:
79/79 [==============================] - 0s 580us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0903
16/16 [==============================] - 0s 1ms/step - loss: 1.8542
16/16 [==============================] - 0s 847us/step - loss: 2.2043
16/16 [==============================] - 0s 1ms/step - loss: 2.2266
16/16 [==============================] - 0s 1ms/step - loss: 2.1917
16/16 [==============================] - 0s 2ms/step - loss: 2.1551
16/16 [==============================] - 0s 838us/step - loss: 2.1206
16/16 [==============================] - 0s 813us/step - loss: 2.1062
16/16 [==============================] - 0s 2ms/step - loss: 2.1015
16/16 [==============================] - 0s 1ms/step - loss: 2.1003

Testing for epoch 39 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0893
16/16 [==============================] - 0s 2ms/step - loss: 1.8766
16/16 [==============================] - 0s 2ms/step - loss: 2.2291
16/16 [==============================] - 0s 1ms/step - loss: 2.2482
16/16 [==============================] - 0s 1ms/step - loss: 2.2101
16/16 [==============================] - 0s 849us/step - loss: 2.1701
16/16 [==============================] - 0s 2ms/step - loss: 2.1335
16/16 [==============================] - 0s 850us/step - loss: 2.1184
16/16 [==============================] - 0s 2ms/step - loss: 2.1136
16/16 [==============================] - 0s 2ms/step - loss: 2.1124

Testing for epoch 39 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 891us/step - loss: 0.0867
16/16 [==============================] - 0s 871us/step - loss: 1.8584
16/16 [==============================] - 0s 2ms/step - loss: 2.2157
16/16 [==============================] - 0s 1ms/step - loss: 2.2382
16/16 [==============================] - 0s 2ms/step - loss: 2.2020
16/16 [==============================] - 0s 901us/step - loss: 2.1634
16/16 [==============================] - 0s 2ms/step - loss: 2.1276
16/16 [==============================] - 0s 2ms/step - loss: 2.1126
16/16 [==============================] - 0s 884us/step - loss: 2.1077
16/16 [==============================] - 0s 2ms/step - loss: 2.1064
Epoch 40 of 60

Testing for epoch 40 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 842us/step - loss: 0.0872
16/16 [==============================] - 0s 844us/step - loss: 1.8698
16/16 [==============================] - 0s 804us/step - loss: 2.2247
16/16 [==============================] - 0s 768us/step - loss: 2.2441
16/16 [==============================] - 0s 1ms/step - loss: 2.2077
16/16 [==============================] - 0s 774us/step - loss: 2.1697
16/16 [==============================] - 0s 2ms/step - loss: 2.1347
16/16 [==============================] - 0s 774us/step - loss: 2.1201
16/16 [==============================] - 0s 2ms/step - loss: 2.1153
16/16 [==============================] - 0s 2ms/step - loss: 2.1141

Testing for epoch 40 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 825us/step - loss: 0.0887
16/16 [==============================] - 0s 2ms/step - loss: 1.9279
16/16 [==============================] - 0s 2ms/step - loss: 2.2914
16/16 [==============================] - 0s 2ms/step - loss: 2.3128
16/16 [==============================] - 0s 877us/step - loss: 2.2753
16/16 [==============================] - 0s 2ms/step - loss: 2.2351
16/16 [==============================] - 0s 2ms/step - loss: 2.1981
16/16 [==============================] - 0s 2ms/step - loss: 2.1826
16/16 [==============================] - 0s 2ms/step - loss: 2.1773
16/16 [==============================] - 0s 2ms/step - loss: 2.1759

Testing for epoch 40 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 979us/step - loss: 0.0863
16/16 [==============================] - 0s 2ms/step - loss: 1.8649
16/16 [==============================] - 0s 882us/step - loss: 2.2107
16/16 [==============================] - 0s 815us/step - loss: 2.2330
16/16 [==============================] - 0s 815us/step - loss: 2.1980
16/16 [==============================] - 0s 1ms/step - loss: 2.1596
16/16 [==============================] - 0s 797us/step - loss: 2.1247
16/16 [==============================] - 0s 799us/step - loss: 2.1099
16/16 [==============================] - 0s 2ms/step - loss: 2.1049
16/16 [==============================] - 0s 1ms/step - loss: 2.1035

Testing for epoch 40 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0867
16/16 [==============================] - 0s 2ms/step - loss: 1.8926
16/16 [==============================] - 0s 2ms/step - loss: 2.2387
16/16 [==============================] - 0s 1ms/step - loss: 2.2572
16/16 [==============================] - 0s 2ms/step - loss: 2.2203
16/16 [==============================] - 0s 1ms/step - loss: 2.1795
16/16 [==============================] - 0s 2ms/step - loss: 2.1427
16/16 [==============================] - 0s 2ms/step - loss: 2.1275
16/16 [==============================] - 0s 815us/step - loss: 2.1227
16/16 [==============================] - 0s 772us/step - loss: 2.1215

Testing for epoch 40 index 5:
79/79 [==============================] - 0s 588us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0845
16/16 [==============================] - 0s 2ms/step - loss: 1.9074
16/16 [==============================] - 0s 2ms/step - loss: 2.2525
16/16 [==============================] - 0s 950us/step - loss: 2.2665
16/16 [==============================] - 0s 1ms/step - loss: 2.2283
16/16 [==============================] - 0s 1ms/step - loss: 2.1883
16/16 [==============================] - 0s 1ms/step - loss: 2.1520
16/16 [==============================] - 0s 2ms/step - loss: 2.1369
16/16 [==============================] - 0s 2ms/step - loss: 2.1319
16/16 [==============================] - 0s 1ms/step - loss: 2.1307
Epoch 41 of 60

Testing for epoch 41 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 950us/step - loss: 0.0841
16/16 [==============================] - 0s 2ms/step - loss: 1.9306
16/16 [==============================] - 0s 984us/step - loss: 2.2909
16/16 [==============================] - 0s 2ms/step - loss: 2.3115
16/16 [==============================] - 0s 2ms/step - loss: 2.2725
16/16 [==============================] - 0s 871us/step - loss: 2.2312
16/16 [==============================] - 0s 2ms/step - loss: 2.1939
16/16 [==============================] - 0s 934us/step - loss: 2.1781
16/16 [==============================] - 0s 2ms/step - loss: 2.1728
16/16 [==============================] - 0s 854us/step - loss: 2.1713

Testing for epoch 41 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 911us/step - loss: 0.0870
16/16 [==============================] - 0s 826us/step - loss: 1.8677
16/16 [==============================] - 0s 2ms/step - loss: 2.2120
16/16 [==============================] - 0s 2ms/step - loss: 2.2279
16/16 [==============================] - 0s 2ms/step - loss: 2.1895
16/16 [==============================] - 0s 926us/step - loss: 2.1490
16/16 [==============================] - 0s 2ms/step - loss: 2.1126
16/16 [==============================] - 0s 2ms/step - loss: 2.0975
16/16 [==============================] - 0s 925us/step - loss: 2.0925
16/16 [==============================] - 0s 2ms/step - loss: 2.0912

Testing for epoch 41 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0831
16/16 [==============================] - 0s 1ms/step - loss: 1.8881
16/16 [==============================] - 0s 836us/step - loss: 2.2354
16/16 [==============================] - 0s 835us/step - loss: 2.2486
16/16 [==============================] - 0s 779us/step - loss: 2.2092
16/16 [==============================] - 0s 834us/step - loss: 2.1679
16/16 [==============================] - 0s 790us/step - loss: 2.1312
16/16 [==============================] - 0s 986us/step - loss: 2.1162
16/16 [==============================] - 0s 2ms/step - loss: 2.1113
16/16 [==============================] - 0s 848us/step - loss: 2.1101

Testing for epoch 41 index 4:
79/79 [==============================] - 0s 979us/step
16/16 [==============================] - 0s 813us/step - loss: 0.0874
16/16 [==============================] - 0s 1ms/step - loss: 1.9374
16/16 [==============================] - 0s 2ms/step - loss: 2.2990
16/16 [==============================] - 0s 2ms/step - loss: 2.3185
16/16 [==============================] - 0s 2ms/step - loss: 2.2794
16/16 [==============================] - 0s 836us/step - loss: 2.2378
16/16 [==============================] - 0s 1ms/step - loss: 2.2000
16/16 [==============================] - 0s 792us/step - loss: 2.1843
16/16 [==============================] - 0s 837us/step - loss: 2.1793
16/16 [==============================] - 0s 819us/step - loss: 2.1781

Testing for epoch 41 index 5:
79/79 [==============================] - 0s 526us/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0842
16/16 [==============================] - 0s 2ms/step - loss: 1.9564
16/16 [==============================] - 0s 2ms/step - loss: 2.3112
16/16 [==============================] - 0s 899us/step - loss: 2.3275
16/16 [==============================] - 0s 791us/step - loss: 2.2860
16/16 [==============================] - 0s 2ms/step - loss: 2.2426
16/16 [==============================] - 0s 1ms/step - loss: 2.2043
16/16 [==============================] - 0s 1ms/step - loss: 2.1885
16/16 [==============================] - 0s 867us/step - loss: 2.1834
16/16 [==============================] - 0s 2ms/step - loss: 2.1821
Epoch 42 of 60

Testing for epoch 42 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0828
16/16 [==============================] - 0s 2ms/step - loss: 1.8963
16/16 [==============================] - 0s 2ms/step - loss: 2.2450
16/16 [==============================] - 0s 2ms/step - loss: 2.2656
16/16 [==============================] - 0s 887us/step - loss: 2.2272
16/16 [==============================] - 0s 838us/step - loss: 2.1849
16/16 [==============================] - 0s 2ms/step - loss: 2.1470
16/16 [==============================] - 0s 799us/step - loss: 2.1315
16/16 [==============================] - 0s 2ms/step - loss: 2.1264
16/16 [==============================] - 0s 804us/step - loss: 2.1250

Testing for epoch 42 index 2:
79/79 [==============================] - 0s 605us/step
16/16 [==============================] - 0s 804us/step - loss: 0.0808
16/16 [==============================] - 0s 773us/step - loss: 1.9638
16/16 [==============================] - 0s 2ms/step - loss: 2.3189
16/16 [==============================] - 0s 2ms/step - loss: 2.3349
16/16 [==============================] - 0s 911us/step - loss: 2.2927
16/16 [==============================] - 0s 2ms/step - loss: 2.2480
16/16 [==============================] - 0s 2ms/step - loss: 2.2085
16/16 [==============================] - 0s 2ms/step - loss: 2.1925
16/16 [==============================] - 0s 2ms/step - loss: 2.1874
16/16 [==============================] - 0s 973us/step - loss: 2.1862

Testing for epoch 42 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0846
16/16 [==============================] - 0s 2ms/step - loss: 1.8955
16/16 [==============================] - 0s 916us/step - loss: 2.2324
16/16 [==============================] - 0s 1ms/step - loss: 2.2467
16/16 [==============================] - 0s 946us/step - loss: 2.2042
16/16 [==============================] - 0s 798us/step - loss: 2.1607
16/16 [==============================] - 0s 790us/step - loss: 2.1227
16/16 [==============================] - 0s 798us/step - loss: 2.1076
16/16 [==============================] - 0s 2ms/step - loss: 2.1027
16/16 [==============================] - 0s 2ms/step - loss: 2.1016

Testing for epoch 42 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 905us/step - loss: 0.0836
16/16 [==============================] - 0s 2ms/step - loss: 1.9991
16/16 [==============================] - 0s 836us/step - loss: 2.3622
16/16 [==============================] - 0s 1ms/step - loss: 2.3816
16/16 [==============================] - 0s 790us/step - loss: 2.3384
16/16 [==============================] - 0s 845us/step - loss: 2.2929
16/16 [==============================] - 0s 1ms/step - loss: 2.2527
16/16 [==============================] - 0s 2ms/step - loss: 2.2364
16/16 [==============================] - 0s 2ms/step - loss: 2.2311
16/16 [==============================] - 0s 2ms/step - loss: 2.2298

Testing for epoch 42 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0845
16/16 [==============================] - 0s 2ms/step - loss: 1.9443
16/16 [==============================] - 0s 906us/step - loss: 2.2978
16/16 [==============================] - 0s 2ms/step - loss: 2.3189
16/16 [==============================] - 0s 2ms/step - loss: 2.2776
16/16 [==============================] - 0s 868us/step - loss: 2.2338
16/16 [==============================] - 0s 783us/step - loss: 2.1949
16/16 [==============================] - 0s 2ms/step - loss: 2.1792
16/16 [==============================] - 0s 875us/step - loss: 2.1741
16/16 [==============================] - 0s 794us/step - loss: 2.1728
Epoch 43 of 60

Testing for epoch 43 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0823
16/16 [==============================] - 0s 927us/step - loss: 1.9151
16/16 [==============================] - 0s 2ms/step - loss: 2.2581
16/16 [==============================] - 0s 784us/step - loss: 2.2767
16/16 [==============================] - 0s 802us/step - loss: 2.2351
16/16 [==============================] - 0s 784us/step - loss: 2.1921
16/16 [==============================] - 0s 787us/step - loss: 2.1545
16/16 [==============================] - 0s 788us/step - loss: 2.1390
16/16 [==============================] - 0s 796us/step - loss: 2.1340
16/16 [==============================] - 0s 813us/step - loss: 2.1327

Testing for epoch 43 index 2:
79/79 [==============================] - 0s 2ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0812
16/16 [==============================] - 0s 826us/step - loss: 1.9665
16/16 [==============================] - 0s 802us/step - loss: 2.3183
16/16 [==============================] - 0s 770us/step - loss: 2.3381
16/16 [==============================] - 0s 886us/step - loss: 2.2949
16/16 [==============================] - 0s 803us/step - loss: 2.2495
16/16 [==============================] - 0s 2ms/step - loss: 2.2099
16/16 [==============================] - 0s 801us/step - loss: 2.1937
16/16 [==============================] - 0s 780us/step - loss: 2.1884
16/16 [==============================] - 0s 796us/step - loss: 2.1870

Testing for epoch 43 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 971us/step - loss: 0.0828
16/16 [==============================] - 0s 1ms/step - loss: 1.9538
16/16 [==============================] - 0s 2ms/step - loss: 2.3013
16/16 [==============================] - 0s 893us/step - loss: 2.3215
16/16 [==============================] - 0s 2ms/step - loss: 2.2813
16/16 [==============================] - 0s 2ms/step - loss: 2.2395
16/16 [==============================] - 0s 2ms/step - loss: 2.2024
16/16 [==============================] - 0s 837us/step - loss: 2.1874
16/16 [==============================] - 0s 965us/step - loss: 2.1825
16/16 [==============================] - 0s 2ms/step - loss: 2.1813

Testing for epoch 43 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0801
16/16 [==============================] - 0s 808us/step - loss: 1.9759
16/16 [==============================] - 0s 796us/step - loss: 2.3167
16/16 [==============================] - 0s 789us/step - loss: 2.3298
16/16 [==============================] - 0s 785us/step - loss: 2.2839
16/16 [==============================] - 0s 797us/step - loss: 2.2360
16/16 [==============================] - 0s 860us/step - loss: 2.1951
16/16 [==============================] - 0s 2ms/step - loss: 2.1788
16/16 [==============================] - 0s 806us/step - loss: 2.1737
16/16 [==============================] - 0s 1ms/step - loss: 2.1725

Testing for epoch 43 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 925us/step - loss: 0.0786
16/16 [==============================] - 0s 1ms/step - loss: 1.9918
16/16 [==============================] - 0s 980us/step - loss: 2.3466
16/16 [==============================] - 0s 1ms/step - loss: 2.3675
16/16 [==============================] - 0s 2ms/step - loss: 2.3251
16/16 [==============================] - 0s 2ms/step - loss: 2.2796
16/16 [==============================] - 0s 859us/step - loss: 2.2397
16/16 [==============================] - 0s 1ms/step - loss: 2.2232
16/16 [==============================] - 0s 1ms/step - loss: 2.2179
16/16 [==============================] - 0s 876us/step - loss: 2.2165
Epoch 44 of 60

Testing for epoch 44 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0815
16/16 [==============================] - 0s 1ms/step - loss: 1.9740
16/16 [==============================] - 0s 922us/step - loss: 2.3169
16/16 [==============================] - 0s 2ms/step - loss: 2.3323
16/16 [==============================] - 0s 2ms/step - loss: 2.2878
16/16 [==============================] - 0s 801us/step - loss: 2.2418
16/16 [==============================] - 0s 900us/step - loss: 2.2026
16/16 [==============================] - 0s 2ms/step - loss: 2.1868
16/16 [==============================] - 0s 810us/step - loss: 2.1818
16/16 [==============================] - 0s 2ms/step - loss: 2.1806

Testing for epoch 44 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0805
16/16 [==============================] - 0s 2ms/step - loss: 2.0101
16/16 [==============================] - 0s 2ms/step - loss: 2.3641
16/16 [==============================] - 0s 892us/step - loss: 2.3810
16/16 [==============================] - 0s 812us/step - loss: 2.3374
16/16 [==============================] - 0s 2ms/step - loss: 2.2919
16/16 [==============================] - 0s 813us/step - loss: 2.2518
16/16 [==============================] - 0s 764us/step - loss: 2.2356
16/16 [==============================] - 0s 806us/step - loss: 2.2304
16/16 [==============================] - 0s 2ms/step - loss: 2.2291

Testing for epoch 44 index 3:
79/79 [==============================] - 0s 534us/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0805
16/16 [==============================] - 0s 2ms/step - loss: 1.9682
16/16 [==============================] - 0s 2ms/step - loss: 2.3102
16/16 [==============================] - 0s 2ms/step - loss: 2.3230
16/16 [==============================] - 0s 2ms/step - loss: 2.2792
16/16 [==============================] - 0s 914us/step - loss: 2.2341
16/16 [==============================] - 0s 807us/step - loss: 2.1954
16/16 [==============================] - 0s 795us/step - loss: 2.1798
16/16 [==============================] - 0s 789us/step - loss: 2.1748
16/16 [==============================] - 0s 793us/step - loss: 2.1735

Testing for epoch 44 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 989us/step - loss: 0.0786
16/16 [==============================] - 0s 2ms/step - loss: 1.9769
16/16 [==============================] - 0s 912us/step - loss: 2.3185
16/16 [==============================] - 0s 821us/step - loss: 2.3281
16/16 [==============================] - 0s 808us/step - loss: 2.2839
16/16 [==============================] - 0s 773us/step - loss: 2.2375
16/16 [==============================] - 0s 810us/step - loss: 2.1975
16/16 [==============================] - 0s 798us/step - loss: 2.1816
16/16 [==============================] - 0s 782us/step - loss: 2.1765
16/16 [==============================] - 0s 787us/step - loss: 2.1752

Testing for epoch 44 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 873us/step - loss: 0.0812
16/16 [==============================] - 0s 874us/step - loss: 1.9463
16/16 [==============================] - 0s 822us/step - loss: 2.2893
16/16 [==============================] - 0s 2ms/step - loss: 2.3024
16/16 [==============================] - 0s 821us/step - loss: 2.2596
16/16 [==============================] - 0s 805us/step - loss: 2.2157
16/16 [==============================] - 0s 801us/step - loss: 2.1778
16/16 [==============================] - 0s 2ms/step - loss: 2.1624
16/16 [==============================] - 0s 2ms/step - loss: 2.1575
16/16 [==============================] - 0s 2ms/step - loss: 2.1563
Epoch 45 of 60

Testing for epoch 45 index 1:
79/79 [==============================] - 0s 884us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0783
16/16 [==============================] - 0s 2ms/step - loss: 2.0198
16/16 [==============================] - 0s 1ms/step - loss: 2.3791
16/16 [==============================] - 0s 1ms/step - loss: 2.3923
16/16 [==============================] - 0s 2ms/step - loss: 2.3469
16/16 [==============================] - 0s 903us/step - loss: 2.2994
16/16 [==============================] - 0s 1ms/step - loss: 2.2585
16/16 [==============================] - 0s 1ms/step - loss: 2.2422
16/16 [==============================] - 0s 814us/step - loss: 2.2371
16/16 [==============================] - 0s 2ms/step - loss: 2.2358

Testing for epoch 45 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 830us/step - loss: 0.0780
16/16 [==============================] - 0s 2ms/step - loss: 1.9802
16/16 [==============================] - 0s 839us/step - loss: 2.3284
16/16 [==============================] - 0s 2ms/step - loss: 2.3400
16/16 [==============================] - 0s 2ms/step - loss: 2.2966
16/16 [==============================] - 0s 2ms/step - loss: 2.2509
16/16 [==============================] - 0s 910us/step - loss: 2.2120
16/16 [==============================] - 0s 2ms/step - loss: 2.1964
16/16 [==============================] - 0s 827us/step - loss: 2.1915
16/16 [==============================] - 0s 788us/step - loss: 2.1903

Testing for epoch 45 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 828us/step - loss: 0.0798
16/16 [==============================] - 0s 865us/step - loss: 1.9756
16/16 [==============================] - 0s 1ms/step - loss: 2.3235
16/16 [==============================] - 0s 819us/step - loss: 2.3351
16/16 [==============================] - 0s 797us/step - loss: 2.2903
16/16 [==============================] - 0s 2ms/step - loss: 2.2432
16/16 [==============================] - 0s 1ms/step - loss: 2.2031
16/16 [==============================] - 0s 801us/step - loss: 2.1873
16/16 [==============================] - 0s 809us/step - loss: 2.1823
16/16 [==============================] - 0s 2ms/step - loss: 2.1811

Testing for epoch 45 index 4:
79/79 [==============================] - 0s 535us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0758
16/16 [==============================] - 0s 1ms/step - loss: 1.9627
16/16 [==============================] - 0s 2ms/step - loss: 2.3096
16/16 [==============================] - 0s 2ms/step - loss: 2.3244
16/16 [==============================] - 0s 1ms/step - loss: 2.2828
16/16 [==============================] - 0s 2ms/step - loss: 2.2391
16/16 [==============================] - 0s 2ms/step - loss: 2.2016
16/16 [==============================] - 0s 2ms/step - loss: 2.1862
16/16 [==============================] - 0s 3ms/step - loss: 2.1811
16/16 [==============================] - 0s 2ms/step - loss: 2.1798

Testing for epoch 45 index 5:
79/79 [==============================] - 0s 688us/step
16/16 [==============================] - 0s 835us/step - loss: 0.0776
16/16 [==============================] - 0s 2ms/step - loss: 2.0054
16/16 [==============================] - 0s 841us/step - loss: 2.3561
16/16 [==============================] - 0s 802us/step - loss: 2.3644
16/16 [==============================] - 0s 2ms/step - loss: 2.3185
16/16 [==============================] - 0s 2ms/step - loss: 2.2706
16/16 [==============================] - 0s 2ms/step - loss: 2.2302
16/16 [==============================] - 0s 2ms/step - loss: 2.2141
16/16 [==============================] - 0s 1ms/step - loss: 2.2091
16/16 [==============================] - 0s 867us/step - loss: 2.2078
Epoch 46 of 60

Testing for epoch 46 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0766
16/16 [==============================] - 0s 910us/step - loss: 2.0468
16/16 [==============================] - 0s 852us/step - loss: 2.4075
16/16 [==============================] - 0s 1ms/step - loss: 2.4124
16/16 [==============================] - 0s 859us/step - loss: 2.3621
16/16 [==============================] - 0s 1ms/step - loss: 2.3090
16/16 [==============================] - 0s 2ms/step - loss: 2.2651
16/16 [==============================] - 0s 867us/step - loss: 2.2476
16/16 [==============================] - 0s 1ms/step - loss: 2.2421
16/16 [==============================] - 0s 819us/step - loss: 2.2408

Testing for epoch 46 index 2:
79/79 [==============================] - 0s 2ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0757
16/16 [==============================] - 0s 874us/step - loss: 2.0077
16/16 [==============================] - 0s 2ms/step - loss: 2.3639
16/16 [==============================] - 0s 2ms/step - loss: 2.3732
16/16 [==============================] - 0s 2ms/step - loss: 2.3268
16/16 [==============================] - 0s 2ms/step - loss: 2.2781
16/16 [==============================] - 0s 2ms/step - loss: 2.2369
16/16 [==============================] - 0s 925us/step - loss: 2.2206
16/16 [==============================] - 0s 865us/step - loss: 2.2156
16/16 [==============================] - 0s 803us/step - loss: 2.2144

Testing for epoch 46 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0762
16/16 [==============================] - 0s 838us/step - loss: 2.0334
16/16 [==============================] - 0s 2ms/step - loss: 2.3980
16/16 [==============================] - 0s 2ms/step - loss: 2.4079
16/16 [==============================] - 0s 1ms/step - loss: 2.3617
16/16 [==============================] - 0s 2ms/step - loss: 2.3134
16/16 [==============================] - 0s 2ms/step - loss: 2.2729
16/16 [==============================] - 0s 2ms/step - loss: 2.2567
16/16 [==============================] - 0s 2ms/step - loss: 2.2516
16/16 [==============================] - 0s 860us/step - loss: 2.2503

Testing for epoch 46 index 4:
79/79 [==============================] - 0s 551us/step
16/16 [==============================] - 0s 994us/step - loss: 0.0765
16/16 [==============================] - 0s 2ms/step - loss: 1.9977
16/16 [==============================] - 0s 1ms/step - loss: 2.3481
16/16 [==============================] - 0s 2ms/step - loss: 2.3571
16/16 [==============================] - 0s 842us/step - loss: 2.3110
16/16 [==============================] - 0s 2ms/step - loss: 2.2619
16/16 [==============================] - 0s 1ms/step - loss: 2.2208
16/16 [==============================] - 0s 2ms/step - loss: 2.2046
16/16 [==============================] - 0s 865us/step - loss: 2.1997
16/16 [==============================] - 0s 1ms/step - loss: 2.1986

Testing for epoch 46 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0730
16/16 [==============================] - 0s 806us/step - loss: 2.0393
16/16 [==============================] - 0s 803us/step - loss: 2.3969
16/16 [==============================] - 0s 2ms/step - loss: 2.4039
16/16 [==============================] - 0s 2ms/step - loss: 2.3575
16/16 [==============================] - 0s 2ms/step - loss: 2.3071
16/16 [==============================] - 0s 2ms/step - loss: 2.2653
16/16 [==============================] - 0s 1ms/step - loss: 2.2487
16/16 [==============================] - 0s 2ms/step - loss: 2.2435
16/16 [==============================] - 0s 918us/step - loss: 2.2423
Epoch 47 of 60

Testing for epoch 47 index 1:
79/79 [==============================] - 0s 981us/step
16/16 [==============================] - 0s 798us/step - loss: 0.0779
16/16 [==============================] - 0s 796us/step - loss: 1.9997
16/16 [==============================] - 0s 815us/step - loss: 2.3521
16/16 [==============================] - 0s 808us/step - loss: 2.3601
16/16 [==============================] - 0s 786us/step - loss: 2.3139
16/16 [==============================] - 0s 2ms/step - loss: 2.2648
16/16 [==============================] - 0s 816us/step - loss: 2.2239
16/16 [==============================] - 0s 2ms/step - loss: 2.2078
16/16 [==============================] - 0s 2ms/step - loss: 2.2027
16/16 [==============================] - 0s 832us/step - loss: 2.2015

Testing for epoch 47 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 912us/step - loss: 0.0745
16/16 [==============================] - 0s 1ms/step - loss: 2.0626
16/16 [==============================] - 0s 1ms/step - loss: 2.4260
16/16 [==============================] - 0s 2ms/step - loss: 2.4307
16/16 [==============================] - 0s 810us/step - loss: 2.3819
16/16 [==============================] - 0s 2ms/step - loss: 2.3300
16/16 [==============================] - 0s 842us/step - loss: 2.2866
16/16 [==============================] - 0s 1ms/step - loss: 2.2695
16/16 [==============================] - 0s 2ms/step - loss: 2.2643
16/16 [==============================] - 0s 844us/step - loss: 2.2632

Testing for epoch 47 index 3:
79/79 [==============================] - 0s 534us/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0758
16/16 [==============================] - 0s 2ms/step - loss: 2.0080
16/16 [==============================] - 0s 816us/step - loss: 2.3568
16/16 [==============================] - 0s 780us/step - loss: 2.3618
16/16 [==============================] - 0s 2ms/step - loss: 2.3151
16/16 [==============================] - 0s 1ms/step - loss: 2.2654
16/16 [==============================] - 0s 801us/step - loss: 2.2251
16/16 [==============================] - 0s 772us/step - loss: 2.2091
16/16 [==============================] - 0s 2ms/step - loss: 2.2042
16/16 [==============================] - 0s 881us/step - loss: 2.2030

Testing for epoch 47 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0747
16/16 [==============================] - 0s 2ms/step - loss: 2.0022
16/16 [==============================] - 0s 2ms/step - loss: 2.3530
16/16 [==============================] - 0s 2ms/step - loss: 2.3562
16/16 [==============================] - 0s 2ms/step - loss: 2.3092
16/16 [==============================] - 0s 1ms/step - loss: 2.2594
16/16 [==============================] - 0s 2ms/step - loss: 2.2185
16/16 [==============================] - 0s 2ms/step - loss: 2.2023
16/16 [==============================] - 0s 1ms/step - loss: 2.1973
16/16 [==============================] - 0s 2ms/step - loss: 2.1961

Testing for epoch 47 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0735
16/16 [==============================] - 0s 2ms/step - loss: 1.9776
16/16 [==============================] - 0s 2ms/step - loss: 2.3273
16/16 [==============================] - 0s 965us/step - loss: 2.3315
16/16 [==============================] - 0s 2ms/step - loss: 2.2852
16/16 [==============================] - 0s 836us/step - loss: 2.2358
16/16 [==============================] - 0s 908us/step - loss: 2.1957
16/16 [==============================] - 0s 1ms/step - loss: 2.1799
16/16 [==============================] - 0s 2ms/step - loss: 2.1750
16/16 [==============================] - 0s 792us/step - loss: 2.1739
Epoch 48 of 60

Testing for epoch 48 index 1:
79/79 [==============================] - 0s 2ms/step
16/16 [==============================] - 0s 858us/step - loss: 0.0767
16/16 [==============================] - 0s 2ms/step - loss: 1.9967
16/16 [==============================] - 0s 2ms/step - loss: 2.3544
16/16 [==============================] - 0s 2ms/step - loss: 2.3631
16/16 [==============================] - 0s 881us/step - loss: 2.3191
16/16 [==============================] - 0s 955us/step - loss: 2.2712
16/16 [==============================] - 0s 777us/step - loss: 2.2313
16/16 [==============================] - 0s 1ms/step - loss: 2.2155
16/16 [==============================] - 0s 2ms/step - loss: 2.2106
16/16 [==============================] - 0s 971us/step - loss: 2.2094

Testing for epoch 48 index 2:
79/79 [==============================] - 0s 544us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0744
16/16 [==============================] - 0s 793us/step - loss: 2.0462
16/16 [==============================] - 0s 2ms/step - loss: 2.4087
16/16 [==============================] - 0s 870us/step - loss: 2.4139
16/16 [==============================] - 0s 917us/step - loss: 2.3674
16/16 [==============================] - 0s 1ms/step - loss: 2.3175
16/16 [==============================] - 0s 2ms/step - loss: 2.2763
16/16 [==============================] - 0s 969us/step - loss: 2.2598
16/16 [==============================] - 0s 922us/step - loss: 2.2546
16/16 [==============================] - 0s 2ms/step - loss: 2.2534

Testing for epoch 48 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0725
16/16 [==============================] - 0s 807us/step - loss: 1.9861
16/16 [==============================] - 0s 774us/step - loss: 2.3321
16/16 [==============================] - 0s 2ms/step - loss: 2.3336
16/16 [==============================] - 0s 1ms/step - loss: 2.2873
16/16 [==============================] - 0s 2ms/step - loss: 2.2385
16/16 [==============================] - 0s 2ms/step - loss: 2.1987
16/16 [==============================] - 0s 2ms/step - loss: 2.1831
16/16 [==============================] - 0s 1ms/step - loss: 2.1782
16/16 [==============================] - 0s 2ms/step - loss: 2.1770

Testing for epoch 48 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0735
16/16 [==============================] - 0s 1ms/step - loss: 2.0634
16/16 [==============================] - 0s 1ms/step - loss: 2.4291
16/16 [==============================] - 0s 832us/step - loss: 2.4289
16/16 [==============================] - 0s 2ms/step - loss: 2.3797
16/16 [==============================] - 0s 2ms/step - loss: 2.3278
16/16 [==============================] - 0s 815us/step - loss: 2.2851
16/16 [==============================] - 0s 945us/step - loss: 2.2683
16/16 [==============================] - 0s 832us/step - loss: 2.2629
16/16 [==============================] - 0s 2ms/step - loss: 2.2616

Testing for epoch 48 index 5:
79/79 [==============================] - 0s 2ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0733
16/16 [==============================] - 0s 2ms/step - loss: 2.0079
16/16 [==============================] - 0s 2ms/step - loss: 2.3607
16/16 [==============================] - 0s 1ms/step - loss: 2.3650
16/16 [==============================] - 0s 1ms/step - loss: 2.3203
16/16 [==============================] - 0s 793us/step - loss: 2.2709
16/16 [==============================] - 0s 804us/step - loss: 2.2304
16/16 [==============================] - 0s 780us/step - loss: 2.2146
16/16 [==============================] - 0s 777us/step - loss: 2.2096
16/16 [==============================] - 0s 784us/step - loss: 2.2084
Epoch 49 of 60

Testing for epoch 49 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 810us/step - loss: 0.0732
16/16 [==============================] - 0s 771us/step - loss: 2.0067
16/16 [==============================] - 0s 803us/step - loss: 2.3527
16/16 [==============================] - 0s 775us/step - loss: 2.3471
16/16 [==============================] - 0s 2ms/step - loss: 2.2989
16/16 [==============================] - 0s 916us/step - loss: 2.2484
16/16 [==============================] - 0s 1ms/step - loss: 2.2073
16/16 [==============================] - 0s 2ms/step - loss: 2.1913
16/16 [==============================] - 0s 811us/step - loss: 2.1863
16/16 [==============================] - 0s 846us/step - loss: 2.1852

Testing for epoch 49 index 2:
79/79 [==============================] - 0s 917us/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0743
16/16 [==============================] - 0s 846us/step - loss: 2.0124
16/16 [==============================] - 0s 932us/step - loss: 2.3659
16/16 [==============================] - 0s 2ms/step - loss: 2.3646
16/16 [==============================] - 0s 853us/step - loss: 2.3179
16/16 [==============================] - 0s 917us/step - loss: 2.2673
16/16 [==============================] - 0s 800us/step - loss: 2.2262
16/16 [==============================] - 0s 801us/step - loss: 2.2100
16/16 [==============================] - 0s 1ms/step - loss: 2.2049
16/16 [==============================] - 0s 2ms/step - loss: 2.2037

Testing for epoch 49 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 883us/step - loss: 0.0698
16/16 [==============================] - 0s 875us/step - loss: 2.0977
16/16 [==============================] - 0s 802us/step - loss: 2.4681
16/16 [==============================] - 0s 2ms/step - loss: 2.4630
16/16 [==============================] - 0s 821us/step - loss: 2.4103
16/16 [==============================] - 0s 2ms/step - loss: 2.3536
16/16 [==============================] - 0s 862us/step - loss: 2.3085
16/16 [==============================] - 0s 802us/step - loss: 2.2909
16/16 [==============================] - 0s 876us/step - loss: 2.2855
16/16 [==============================] - 0s 1ms/step - loss: 2.2842

Testing for epoch 49 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 825us/step - loss: 0.0715
16/16 [==============================] - 0s 797us/step - loss: 2.0768
16/16 [==============================] - 0s 788us/step - loss: 2.4382
16/16 [==============================] - 0s 812us/step - loss: 2.4322
16/16 [==============================] - 0s 797us/step - loss: 2.3825
16/16 [==============================] - 0s 819us/step - loss: 2.3286
16/16 [==============================] - 0s 2ms/step - loss: 2.2850
16/16 [==============================] - 0s 2ms/step - loss: 2.2680
16/16 [==============================] - 0s 956us/step - loss: 2.2627
16/16 [==============================] - 0s 1ms/step - loss: 2.2615

Testing for epoch 49 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 883us/step - loss: 0.0726
16/16 [==============================] - 0s 1ms/step - loss: 1.9929
16/16 [==============================] - 0s 862us/step - loss: 2.3413
16/16 [==============================] - 0s 2ms/step - loss: 2.3376
16/16 [==============================] - 0s 813us/step - loss: 2.2908
16/16 [==============================] - 0s 797us/step - loss: 2.2405
16/16 [==============================] - 0s 2ms/step - loss: 2.1999
16/16 [==============================] - 0s 820us/step - loss: 2.1841
16/16 [==============================] - 0s 785us/step - loss: 2.1791
16/16 [==============================] - 0s 1ms/step - loss: 2.1779
Epoch 50 of 60

Testing for epoch 50 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 984us/step - loss: 0.0698
16/16 [==============================] - 0s 879us/step - loss: 2.0829
16/16 [==============================] - 0s 841us/step - loss: 2.4574
16/16 [==============================] - 0s 2ms/step - loss: 2.4604
16/16 [==============================] - 0s 848us/step - loss: 2.4129
16/16 [==============================] - 0s 1ms/step - loss: 2.3622
16/16 [==============================] - 0s 836us/step - loss: 2.3207
16/16 [==============================] - 0s 1ms/step - loss: 2.3042
16/16 [==============================] - 0s 2ms/step - loss: 2.2990
16/16 [==============================] - 0s 849us/step - loss: 2.2977

Testing for epoch 50 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0696
16/16 [==============================] - 0s 2ms/step - loss: 2.0772
16/16 [==============================] - 0s 2ms/step - loss: 2.4379
16/16 [==============================] - 0s 1ms/step - loss: 2.4321
16/16 [==============================] - 0s 911us/step - loss: 2.3819
16/16 [==============================] - 0s 858us/step - loss: 2.3297
16/16 [==============================] - 0s 814us/step - loss: 2.2877
16/16 [==============================] - 0s 834us/step - loss: 2.2713
16/16 [==============================] - 0s 1ms/step - loss: 2.2663
16/16 [==============================] - 0s 2ms/step - loss: 2.2652

Testing for epoch 50 index 3:
79/79 [==============================] - 0s 636us/step
16/16 [==============================] - 0s 801us/step - loss: 0.0736
16/16 [==============================] - 0s 798us/step - loss: 2.0414
16/16 [==============================] - 0s 2ms/step - loss: 2.3966
16/16 [==============================] - 0s 2ms/step - loss: 2.3948
16/16 [==============================] - 0s 2ms/step - loss: 2.3477
16/16 [==============================] - 0s 2ms/step - loss: 2.2969
16/16 [==============================] - 0s 928us/step - loss: 2.2558
16/16 [==============================] - 0s 2ms/step - loss: 2.2397
16/16 [==============================] - 0s 849us/step - loss: 2.2347
16/16 [==============================] - 0s 829us/step - loss: 2.2335

Testing for epoch 50 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0732
16/16 [==============================] - 0s 2ms/step - loss: 2.0117
16/16 [==============================] - 0s 851us/step - loss: 2.3635
16/16 [==============================] - 0s 2ms/step - loss: 2.3617
16/16 [==============================] - 0s 808us/step - loss: 2.3142
16/16 [==============================] - 0s 1ms/step - loss: 2.2635
16/16 [==============================] - 0s 2ms/step - loss: 2.2230
16/16 [==============================] - 0s 935us/step - loss: 2.2070
16/16 [==============================] - 0s 812us/step - loss: 2.2018
16/16 [==============================] - 0s 1ms/step - loss: 2.2004

Testing for epoch 50 index 5:
79/79 [==============================] - 0s 560us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0689
16/16 [==============================] - 0s 2ms/step - loss: 2.0838
16/16 [==============================] - 0s 832us/step - loss: 2.4465
16/16 [==============================] - 0s 805us/step - loss: 2.4418
16/16 [==============================] - 0s 790us/step - loss: 2.3901
16/16 [==============================] - 0s 1ms/step - loss: 2.3353
16/16 [==============================] - 0s 2ms/step - loss: 2.2913
16/16 [==============================] - 0s 816us/step - loss: 2.2745
16/16 [==============================] - 0s 2ms/step - loss: 2.2694
16/16 [==============================] - 0s 2ms/step - loss: 2.2682
Epoch 51 of 60

Testing for epoch 51 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 785us/step - loss: 0.0731
16/16 [==============================] - 0s 2ms/step - loss: 2.0468
16/16 [==============================] - 0s 1ms/step - loss: 2.4122
16/16 [==============================] - 0s 2ms/step - loss: 2.4120
16/16 [==============================] - 0s 832us/step - loss: 2.3644
16/16 [==============================] - 0s 792us/step - loss: 2.3140
16/16 [==============================] - 0s 1ms/step - loss: 2.2734
16/16 [==============================] - 0s 2ms/step - loss: 2.2577
16/16 [==============================] - 0s 2ms/step - loss: 2.2528
16/16 [==============================] - 0s 875us/step - loss: 2.2516

Testing for epoch 51 index 2:
79/79 [==============================] - 0s 996us/step
16/16 [==============================] - 0s 822us/step - loss: 0.0697
16/16 [==============================] - 0s 788us/step - loss: 2.0365
16/16 [==============================] - 0s 2ms/step - loss: 2.3938
16/16 [==============================] - 0s 2ms/step - loss: 2.3895
16/16 [==============================] - 0s 814us/step - loss: 2.3405
16/16 [==============================] - 0s 1ms/step - loss: 2.2889
16/16 [==============================] - 0s 796us/step - loss: 2.2475
16/16 [==============================] - 0s 2ms/step - loss: 2.2314
16/16 [==============================] - 0s 812us/step - loss: 2.2264
16/16 [==============================] - 0s 767us/step - loss: 2.2251

Testing for epoch 51 index 3:
79/79 [==============================] - 0s 994us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0681
16/16 [==============================] - 0s 897us/step - loss: 2.0709
16/16 [==============================] - 0s 2ms/step - loss: 2.4350
16/16 [==============================] - 0s 872us/step - loss: 2.4303
16/16 [==============================] - 0s 2ms/step - loss: 2.3793
16/16 [==============================] - 0s 2ms/step - loss: 2.3248
16/16 [==============================] - 0s 2ms/step - loss: 2.2811
16/16 [==============================] - 0s 906us/step - loss: 2.2643
16/16 [==============================] - 0s 2ms/step - loss: 2.2593
16/16 [==============================] - 0s 2ms/step - loss: 2.2581

Testing for epoch 51 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0708
16/16 [==============================] - 0s 2ms/step - loss: 2.0492
16/16 [==============================] - 0s 2ms/step - loss: 2.4075
16/16 [==============================] - 0s 2ms/step - loss: 2.4020
16/16 [==============================] - 0s 2ms/step - loss: 2.3530
16/16 [==============================] - 0s 889us/step - loss: 2.3003
16/16 [==============================] - 0s 848us/step - loss: 2.2581
16/16 [==============================] - 0s 784us/step - loss: 2.2419
16/16 [==============================] - 0s 2ms/step - loss: 2.2368
16/16 [==============================] - 0s 2ms/step - loss: 2.2356

Testing for epoch 51 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0692
16/16 [==============================] - 0s 2ms/step - loss: 2.1075
16/16 [==============================] - 0s 2ms/step - loss: 2.4767
16/16 [==============================] - 0s 2ms/step - loss: 2.4664
16/16 [==============================] - 0s 2ms/step - loss: 2.4114
16/16 [==============================] - 0s 899us/step - loss: 2.3533
16/16 [==============================] - 0s 948us/step - loss: 2.3075
16/16 [==============================] - 0s 2ms/step - loss: 2.2900
16/16 [==============================] - 0s 924us/step - loss: 2.2846
16/16 [==============================] - 0s 2ms/step - loss: 2.2834
Epoch 52 of 60

Testing for epoch 52 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0685
16/16 [==============================] - 0s 2ms/step - loss: 2.1059
16/16 [==============================] - 0s 2ms/step - loss: 2.4929
16/16 [==============================] - 0s 2ms/step - loss: 2.4970
16/16 [==============================] - 0s 967us/step - loss: 2.4491
16/16 [==============================] - 0s 2ms/step - loss: 2.3965
16/16 [==============================] - 0s 1ms/step - loss: 2.3534
16/16 [==============================] - 0s 848us/step - loss: 2.3367
16/16 [==============================] - 0s 2ms/step - loss: 2.3315
16/16 [==============================] - 0s 825us/step - loss: 2.3301

Testing for epoch 52 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 826us/step - loss: 0.0681
16/16 [==============================] - 0s 826us/step - loss: 2.0712
16/16 [==============================] - 0s 2ms/step - loss: 2.4428
16/16 [==============================] - 0s 2ms/step - loss: 2.4404
16/16 [==============================] - 0s 2ms/step - loss: 2.3914
16/16 [==============================] - 0s 820us/step - loss: 2.3382
16/16 [==============================] - 0s 796us/step - loss: 2.2957
16/16 [==============================] - 0s 1ms/step - loss: 2.2793
16/16 [==============================] - 0s 839us/step - loss: 2.2742
16/16 [==============================] - 0s 2ms/step - loss: 2.2729

Testing for epoch 52 index 3:
79/79 [==============================] - 0s 546us/step
16/16 [==============================] - 0s 788us/step - loss: 0.0695
16/16 [==============================] - 0s 774us/step - loss: 2.0962
16/16 [==============================] - 0s 917us/step - loss: 2.4675
16/16 [==============================] - 0s 2ms/step - loss: 2.4621
16/16 [==============================] - 0s 2ms/step - loss: 2.4117
16/16 [==============================] - 0s 2ms/step - loss: 2.3581
16/16 [==============================] - 0s 980us/step - loss: 2.3151
16/16 [==============================] - 0s 2ms/step - loss: 2.2985
16/16 [==============================] - 0s 2ms/step - loss: 2.2934
16/16 [==============================] - 0s 1ms/step - loss: 2.2921

Testing for epoch 52 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0678
16/16 [==============================] - 0s 2ms/step - loss: 2.0956
16/16 [==============================] - 0s 930us/step - loss: 2.4642
16/16 [==============================] - 0s 915us/step - loss: 2.4530
16/16 [==============================] - 0s 809us/step - loss: 2.3982
16/16 [==============================] - 0s 814us/step - loss: 2.3407
16/16 [==============================] - 0s 808us/step - loss: 2.2950
16/16 [==============================] - 0s 857us/step - loss: 2.2774
16/16 [==============================] - 0s 815us/step - loss: 2.2720
16/16 [==============================] - 0s 2ms/step - loss: 2.2708

Testing for epoch 52 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0698
16/16 [==============================] - 0s 2ms/step - loss: 2.1065
16/16 [==============================] - 0s 1ms/step - loss: 2.4821
16/16 [==============================] - 0s 2ms/step - loss: 2.4692
16/16 [==============================] - 0s 876us/step - loss: 2.4147
16/16 [==============================] - 0s 2ms/step - loss: 2.3576
16/16 [==============================] - 0s 1ms/step - loss: 2.3128
16/16 [==============================] - 0s 872us/step - loss: 2.2956
16/16 [==============================] - 0s 795us/step - loss: 2.2903
16/16 [==============================] - 0s 799us/step - loss: 2.2891
Epoch 53 of 60

Testing for epoch 53 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 902us/step - loss: 0.0672
16/16 [==============================] - 0s 2ms/step - loss: 2.1232
16/16 [==============================] - 0s 2ms/step - loss: 2.5020
16/16 [==============================] - 0s 2ms/step - loss: 2.4944
16/16 [==============================] - 0s 2ms/step - loss: 2.4420
16/16 [==============================] - 0s 2ms/step - loss: 2.3869
16/16 [==============================] - 0s 925us/step - loss: 2.3429
16/16 [==============================] - 0s 836us/step - loss: 2.3260
16/16 [==============================] - 0s 803us/step - loss: 2.3208
16/16 [==============================] - 0s 989us/step - loss: 2.3195

Testing for epoch 53 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 837us/step - loss: 0.0671
16/16 [==============================] - 0s 1ms/step - loss: 2.1642
16/16 [==============================] - 0s 2ms/step - loss: 2.5565
16/16 [==============================] - 0s 1ms/step - loss: 2.5513
16/16 [==============================] - 0s 787us/step - loss: 2.4989
16/16 [==============================] - 0s 851us/step - loss: 2.4426
16/16 [==============================] - 0s 2ms/step - loss: 2.3977
16/16 [==============================] - 0s 2ms/step - loss: 2.3802
16/16 [==============================] - 0s 2ms/step - loss: 2.3747
16/16 [==============================] - 0s 2ms/step - loss: 2.3733

Testing for epoch 53 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0691
16/16 [==============================] - 0s 1ms/step - loss: 2.1175
16/16 [==============================] - 0s 926us/step - loss: 2.4916
16/16 [==============================] - 0s 2ms/step - loss: 2.4838
16/16 [==============================] - 0s 1ms/step - loss: 2.4293
16/16 [==============================] - 0s 2ms/step - loss: 2.3711
16/16 [==============================] - 0s 927us/step - loss: 2.3259
16/16 [==============================] - 0s 2ms/step - loss: 2.3088
16/16 [==============================] - 0s 947us/step - loss: 2.3035
16/16 [==============================] - 0s 1ms/step - loss: 2.3022

Testing for epoch 53 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0647
16/16 [==============================] - 0s 851us/step - loss: 2.1075
16/16 [==============================] - 0s 822us/step - loss: 2.4701
16/16 [==============================] - 0s 832us/step - loss: 2.4603
16/16 [==============================] - 0s 786us/step - loss: 2.4066
16/16 [==============================] - 0s 802us/step - loss: 2.3501
16/16 [==============================] - 0s 779us/step - loss: 2.3057
16/16 [==============================] - 0s 2ms/step - loss: 2.2888
16/16 [==============================] - 0s 1ms/step - loss: 2.2836
16/16 [==============================] - 0s 859us/step - loss: 2.2823

Testing for epoch 53 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0668
16/16 [==============================] - 0s 858us/step - loss: 2.1158
16/16 [==============================] - 0s 2ms/step - loss: 2.4846
16/16 [==============================] - 0s 790us/step - loss: 2.4771
16/16 [==============================] - 0s 779us/step - loss: 2.4247
16/16 [==============================] - 0s 784us/step - loss: 2.3707
16/16 [==============================] - 0s 2ms/step - loss: 2.3276
16/16 [==============================] - 0s 2ms/step - loss: 2.3107
16/16 [==============================] - 0s 2ms/step - loss: 2.3053
16/16 [==============================] - 0s 2ms/step - loss: 2.3039
Epoch 54 of 60

Testing for epoch 54 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0669
16/16 [==============================] - 0s 1ms/step - loss: 2.1070
16/16 [==============================] - 0s 958us/step - loss: 2.4657
16/16 [==============================] - 0s 861us/step - loss: 2.4552
16/16 [==============================] - 0s 1ms/step - loss: 2.4006
16/16 [==============================] - 0s 824us/step - loss: 2.3447
16/16 [==============================] - 0s 784us/step - loss: 2.3012
16/16 [==============================] - 0s 2ms/step - loss: 2.2846
16/16 [==============================] - 0s 1ms/step - loss: 2.2795
16/16 [==============================] - 0s 2ms/step - loss: 2.2784

Testing for epoch 54 index 2:
79/79 [==============================] - 0s 664us/step
16/16 [==============================] - 0s 792us/step - loss: 0.0690
16/16 [==============================] - 0s 2ms/step - loss: 2.1828
16/16 [==============================] - 0s 2ms/step - loss: 2.5562
16/16 [==============================] - 0s 2ms/step - loss: 2.5489
16/16 [==============================] - 0s 1ms/step - loss: 2.4953
16/16 [==============================] - 0s 828us/step - loss: 2.4394
16/16 [==============================] - 0s 777us/step - loss: 2.3954
16/16 [==============================] - 0s 2ms/step - loss: 2.3785
16/16 [==============================] - 0s 1ms/step - loss: 2.3734
16/16 [==============================] - 0s 819us/step - loss: 2.3721

Testing for epoch 54 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0674
16/16 [==============================] - 0s 934us/step - loss: 2.0479
16/16 [==============================] - 0s 2ms/step - loss: 2.3901
16/16 [==============================] - 0s 904us/step - loss: 2.3841
16/16 [==============================] - 0s 799us/step - loss: 2.3349
16/16 [==============================] - 0s 817us/step - loss: 2.2833
16/16 [==============================] - 0s 799us/step - loss: 2.2424
16/16 [==============================] - 0s 812us/step - loss: 2.2267
16/16 [==============================] - 0s 778us/step - loss: 2.2218
16/16 [==============================] - 0s 1ms/step - loss: 2.2207

Testing for epoch 54 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0671
16/16 [==============================] - 0s 880us/step - loss: 2.1221
16/16 [==============================] - 0s 2ms/step - loss: 2.4701
16/16 [==============================] - 0s 2ms/step - loss: 2.4558
16/16 [==============================] - 0s 878us/step - loss: 2.4003
16/16 [==============================] - 0s 812us/step - loss: 2.3425
16/16 [==============================] - 0s 788us/step - loss: 2.2978
16/16 [==============================] - 0s 814us/step - loss: 2.2808
16/16 [==============================] - 0s 796us/step - loss: 2.2756
16/16 [==============================] - 0s 898us/step - loss: 2.2743

Testing for epoch 54 index 5:
79/79 [==============================] - 0s 2ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0663
16/16 [==============================] - 0s 951us/step - loss: 2.1344
16/16 [==============================] - 0s 2ms/step - loss: 2.4931
16/16 [==============================] - 0s 2ms/step - loss: 2.4845
16/16 [==============================] - 0s 856us/step - loss: 2.4344
16/16 [==============================] - 0s 2ms/step - loss: 2.3803
16/16 [==============================] - 0s 826us/step - loss: 2.3370
16/16 [==============================] - 0s 805us/step - loss: 2.3205
16/16 [==============================] - 0s 2ms/step - loss: 2.3155
16/16 [==============================] - 0s 1ms/step - loss: 2.3143
Epoch 55 of 60

Testing for epoch 55 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0679
16/16 [==============================] - 0s 836us/step - loss: 2.0815
16/16 [==============================] - 0s 1ms/step - loss: 2.4278
16/16 [==============================] - 0s 789us/step - loss: 2.4172
16/16 [==============================] - 0s 746us/step - loss: 2.3647
16/16 [==============================] - 0s 769us/step - loss: 2.3094
16/16 [==============================] - 0s 869us/step - loss: 2.2667
16/16 [==============================] - 0s 2ms/step - loss: 2.2502
16/16 [==============================] - 0s 2ms/step - loss: 2.2451
16/16 [==============================] - 0s 2ms/step - loss: 2.2437

Testing for epoch 55 index 2:
79/79 [==============================] - 0s 935us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0648
16/16 [==============================] - 0s 846us/step - loss: 2.1377
16/16 [==============================] - 0s 818us/step - loss: 2.4996
16/16 [==============================] - 0s 802us/step - loss: 2.4929
16/16 [==============================] - 0s 797us/step - loss: 2.4417
16/16 [==============================] - 0s 2ms/step - loss: 2.3863
16/16 [==============================] - 0s 2ms/step - loss: 2.3429
16/16 [==============================] - 0s 2ms/step - loss: 2.3263
16/16 [==============================] - 0s 2ms/step - loss: 2.3214
16/16 [==============================] - 0s 927us/step - loss: 2.3203

Testing for epoch 55 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0655
16/16 [==============================] - 0s 1ms/step - loss: 2.1209
16/16 [==============================] - 0s 897us/step - loss: 2.4647
16/16 [==============================] - 0s 1ms/step - loss: 2.4525
16/16 [==============================] - 0s 893us/step - loss: 2.4002
16/16 [==============================] - 0s 2ms/step - loss: 2.3439
16/16 [==============================] - 0s 841us/step - loss: 2.2999
16/16 [==============================] - 0s 2ms/step - loss: 2.2832
16/16 [==============================] - 0s 842us/step - loss: 2.2781
16/16 [==============================] - 0s 795us/step - loss: 2.2769

Testing for epoch 55 index 4:
79/79 [==============================] - 0s 536us/step
16/16 [==============================] - 0s 790us/step - loss: 0.0656
16/16 [==============================] - 0s 772us/step - loss: 2.1164
16/16 [==============================] - 0s 884us/step - loss: 2.4588
16/16 [==============================] - 0s 1ms/step - loss: 2.4436
16/16 [==============================] - 0s 2ms/step - loss: 2.3897
16/16 [==============================] - 0s 876us/step - loss: 2.3342
16/16 [==============================] - 0s 1ms/step - loss: 2.2912
16/16 [==============================] - 0s 968us/step - loss: 2.2750
16/16 [==============================] - 0s 2ms/step - loss: 2.2701
16/16 [==============================] - 0s 2ms/step - loss: 2.2690

Testing for epoch 55 index 5:
79/79 [==============================] - 0s 820us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0652
16/16 [==============================] - 0s 2ms/step - loss: 2.1518
16/16 [==============================] - 0s 2ms/step - loss: 2.5130
16/16 [==============================] - 0s 2ms/step - loss: 2.4994
16/16 [==============================] - 0s 1ms/step - loss: 2.4455
16/16 [==============================] - 0s 930us/step - loss: 2.3889
16/16 [==============================] - 0s 1ms/step - loss: 2.3445
16/16 [==============================] - 0s 2ms/step - loss: 2.3273
16/16 [==============================] - 0s 2ms/step - loss: 2.3221
16/16 [==============================] - 0s 798us/step - loss: 2.3208
Epoch 56 of 60

Testing for epoch 56 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0643
16/16 [==============================] - 0s 2ms/step - loss: 2.1867
16/16 [==============================] - 0s 1ms/step - loss: 2.5477
16/16 [==============================] - 0s 2ms/step - loss: 2.5307
16/16 [==============================] - 0s 2ms/step - loss: 2.4748
16/16 [==============================] - 0s 846us/step - loss: 2.4170
16/16 [==============================] - 0s 2ms/step - loss: 2.3722
16/16 [==============================] - 0s 892us/step - loss: 2.3551
16/16 [==============================] - 0s 816us/step - loss: 2.3498
16/16 [==============================] - 0s 1ms/step - loss: 2.3486

Testing for epoch 56 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 900us/step - loss: 0.0658
16/16 [==============================] - 0s 2ms/step - loss: 2.1272
16/16 [==============================] - 0s 912us/step - loss: 2.4774
16/16 [==============================] - 0s 2ms/step - loss: 2.4623
16/16 [==============================] - 0s 2ms/step - loss: 2.4081
16/16 [==============================] - 0s 1ms/step - loss: 2.3514
16/16 [==============================] - 0s 813us/step - loss: 2.3074
16/16 [==============================] - 0s 791us/step - loss: 2.2906
16/16 [==============================] - 0s 2ms/step - loss: 2.2855
16/16 [==============================] - 0s 2ms/step - loss: 2.2843

Testing for epoch 56 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0636
16/16 [==============================] - 0s 1ms/step - loss: 2.1518
16/16 [==============================] - 0s 1ms/step - loss: 2.4905
16/16 [==============================] - 0s 825us/step - loss: 2.4634
16/16 [==============================] - 0s 2ms/step - loss: 2.4054
16/16 [==============================] - 0s 1ms/step - loss: 2.3458
16/16 [==============================] - 0s 793us/step - loss: 2.3005
16/16 [==============================] - 0s 820us/step - loss: 2.2834
16/16 [==============================] - 0s 2ms/step - loss: 2.2783
16/16 [==============================] - 0s 2ms/step - loss: 2.2771

Testing for epoch 56 index 4:
79/79 [==============================] - 0s 936us/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0625
16/16 [==============================] - 0s 826us/step - loss: 2.2182
16/16 [==============================] - 0s 828us/step - loss: 2.5822
16/16 [==============================] - 0s 838us/step - loss: 2.5602
16/16 [==============================] - 0s 794us/step - loss: 2.5020
16/16 [==============================] - 0s 857us/step - loss: 2.4396
16/16 [==============================] - 0s 854us/step - loss: 2.3916
16/16 [==============================] - 0s 831us/step - loss: 2.3735
16/16 [==============================] - 0s 831us/step - loss: 2.3679
16/16 [==============================] - 0s 2ms/step - loss: 2.3666

Testing for epoch 56 index 5:
79/79 [==============================] - 0s 2ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0646
16/16 [==============================] - 0s 1ms/step - loss: 2.1041
16/16 [==============================] - 0s 823us/step - loss: 2.4493
16/16 [==============================] - 0s 2ms/step - loss: 2.4361
16/16 [==============================] - 0s 2ms/step - loss: 2.3867
16/16 [==============================] - 0s 848us/step - loss: 2.3334
16/16 [==============================] - 0s 807us/step - loss: 2.2913
16/16 [==============================] - 0s 773us/step - loss: 2.2751
16/16 [==============================] - 0s 1ms/step - loss: 2.2701
16/16 [==============================] - 0s 802us/step - loss: 2.2689
Epoch 57 of 60

Testing for epoch 57 index 1:
79/79 [==============================] - 0s 563us/step
16/16 [==============================] - 0s 775us/step - loss: 0.0614
16/16 [==============================] - 0s 778us/step - loss: 2.1879
16/16 [==============================] - 0s 790us/step - loss: 2.5443
16/16 [==============================] - 0s 774us/step - loss: 2.5183
16/16 [==============================] - 0s 763us/step - loss: 2.4600
16/16 [==============================] - 0s 803us/step - loss: 2.3990
16/16 [==============================] - 0s 776us/step - loss: 2.3521
16/16 [==============================] - 0s 2ms/step - loss: 2.3345
16/16 [==============================] - 0s 2ms/step - loss: 2.3293
16/16 [==============================] - 0s 2ms/step - loss: 2.3280

Testing for epoch 57 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0642
16/16 [==============================] - 0s 2ms/step - loss: 2.1720
16/16 [==============================] - 0s 1ms/step - loss: 2.5330
16/16 [==============================] - 0s 823us/step - loss: 2.5108
16/16 [==============================] - 0s 2ms/step - loss: 2.4543
16/16 [==============================] - 0s 1ms/step - loss: 2.3953
16/16 [==============================] - 0s 763us/step - loss: 2.3507
16/16 [==============================] - 0s 1ms/step - loss: 2.3336
16/16 [==============================] - 0s 2ms/step - loss: 2.3284
16/16 [==============================] - 0s 1ms/step - loss: 2.3272

Testing for epoch 57 index 3:
79/79 [==============================] - 0s 571us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0665
16/16 [==============================] - 0s 847us/step - loss: 2.1306
16/16 [==============================] - 0s 793us/step - loss: 2.4777
16/16 [==============================] - 0s 810us/step - loss: 2.4554
16/16 [==============================] - 0s 2ms/step - loss: 2.4022
16/16 [==============================] - 0s 816us/step - loss: 2.3453
16/16 [==============================] - 0s 2ms/step - loss: 2.3014
16/16 [==============================] - 0s 853us/step - loss: 2.2845
16/16 [==============================] - 0s 2ms/step - loss: 2.2792
16/16 [==============================] - 0s 852us/step - loss: 2.2779

Testing for epoch 57 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 905us/step - loss: 0.0622
16/16 [==============================] - 0s 1ms/step - loss: 2.1649
16/16 [==============================] - 0s 2ms/step - loss: 2.5212
16/16 [==============================] - 0s 2ms/step - loss: 2.4932
16/16 [==============================] - 0s 2ms/step - loss: 2.4352
16/16 [==============================] - 0s 2ms/step - loss: 2.3743
16/16 [==============================] - 0s 2ms/step - loss: 2.3291
16/16 [==============================] - 0s 2ms/step - loss: 2.3120
16/16 [==============================] - 0s 1ms/step - loss: 2.3068
16/16 [==============================] - 0s 873us/step - loss: 2.3056

Testing for epoch 57 index 5:
79/79 [==============================] - 0s 876us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0616
16/16 [==============================] - 0s 768us/step - loss: 2.2366
16/16 [==============================] - 0s 1ms/step - loss: 2.6058
16/16 [==============================] - 0s 2ms/step - loss: 2.5771
16/16 [==============================] - 0s 794us/step - loss: 2.5177
16/16 [==============================] - 0s 785us/step - loss: 2.4556
16/16 [==============================] - 0s 773us/step - loss: 2.4087
16/16 [==============================] - 0s 812us/step - loss: 2.3907
16/16 [==============================] - 0s 784us/step - loss: 2.3851
16/16 [==============================] - 0s 2ms/step - loss: 2.3837
Epoch 58 of 60

Testing for epoch 58 index 1:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0639
16/16 [==============================] - 0s 1ms/step - loss: 2.1615
16/16 [==============================] - 0s 2ms/step - loss: 2.5153
16/16 [==============================] - 0s 895us/step - loss: 2.4855
16/16 [==============================] - 0s 952us/step - loss: 2.4284
16/16 [==============================] - 0s 2ms/step - loss: 2.3688
16/16 [==============================] - 0s 2ms/step - loss: 2.3240
16/16 [==============================] - 0s 1ms/step - loss: 2.3071
16/16 [==============================] - 0s 2ms/step - loss: 2.3021
16/16 [==============================] - 0s 2ms/step - loss: 2.3009

Testing for epoch 58 index 2:
79/79 [==============================] - 0s 547us/step
16/16 [==============================] - 0s 790us/step - loss: 0.0640
16/16 [==============================] - 0s 2ms/step - loss: 2.1783
16/16 [==============================] - 0s 2ms/step - loss: 2.5323
16/16 [==============================] - 0s 2ms/step - loss: 2.5076
16/16 [==============================] - 0s 2ms/step - loss: 2.4511
16/16 [==============================] - 0s 914us/step - loss: 2.3926
16/16 [==============================] - 0s 2ms/step - loss: 2.3478
16/16 [==============================] - 0s 917us/step - loss: 2.3309
16/16 [==============================] - 0s 831us/step - loss: 2.3259
16/16 [==============================] - 0s 819us/step - loss: 2.3248

Testing for epoch 58 index 3:
79/79 [==============================] - 0s 534us/step
16/16 [==============================] - 0s 804us/step - loss: 0.0641
16/16 [==============================] - 0s 2ms/step - loss: 2.2082
16/16 [==============================] - 0s 968us/step - loss: 2.5695
16/16 [==============================] - 0s 869us/step - loss: 2.5436
16/16 [==============================] - 0s 2ms/step - loss: 2.4863
16/16 [==============================] - 0s 2ms/step - loss: 2.4256
16/16 [==============================] - 0s 791us/step - loss: 2.3792
16/16 [==============================] - 0s 1ms/step - loss: 2.3618
16/16 [==============================] - 0s 914us/step - loss: 2.3566
16/16 [==============================] - 0s 796us/step - loss: 2.3554

Testing for epoch 58 index 4:
79/79 [==============================] - 0s 828us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0626
16/16 [==============================] - 0s 2ms/step - loss: 2.1917
16/16 [==============================] - 0s 892us/step - loss: 2.5341
16/16 [==============================] - 0s 808us/step - loss: 2.5009
16/16 [==============================] - 0s 853us/step - loss: 2.4437
16/16 [==============================] - 0s 833us/step - loss: 2.3835
16/16 [==============================] - 0s 803us/step - loss: 2.3382
16/16 [==============================] - 0s 810us/step - loss: 2.3212
16/16 [==============================] - 0s 2ms/step - loss: 2.3161
16/16 [==============================] - 0s 1ms/step - loss: 2.3150

Testing for epoch 58 index 5:
79/79 [==============================] - 0s 966us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0623
16/16 [==============================] - 0s 885us/step - loss: 2.1952
16/16 [==============================] - 0s 2ms/step - loss: 2.5553
16/16 [==============================] - 0s 816us/step - loss: 2.5323
16/16 [==============================] - 0s 785us/step - loss: 2.4776
16/16 [==============================] - 0s 800us/step - loss: 2.4194
16/16 [==============================] - 0s 965us/step - loss: 2.3746
16/16 [==============================] - 0s 2ms/step - loss: 2.3576
16/16 [==============================] - 0s 2ms/step - loss: 2.3524
16/16 [==============================] - 0s 839us/step - loss: 2.3512
Epoch 59 of 60

Testing for epoch 59 index 1:
79/79 [==============================] - 0s 983us/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0624
16/16 [==============================] - 0s 2ms/step - loss: 2.2209
16/16 [==============================] - 0s 2ms/step - loss: 2.5878
16/16 [==============================] - 0s 825us/step - loss: 2.5638
16/16 [==============================] - 0s 2ms/step - loss: 2.5067
16/16 [==============================] - 0s 2ms/step - loss: 2.4469
16/16 [==============================] - 0s 2ms/step - loss: 2.4021
16/16 [==============================] - 0s 2ms/step - loss: 2.3849
16/16 [==============================] - 0s 2ms/step - loss: 2.3796
16/16 [==============================] - 0s 1ms/step - loss: 2.3783

Testing for epoch 59 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0643
16/16 [==============================] - 0s 911us/step - loss: 2.1610
16/16 [==============================] - 0s 2ms/step - loss: 2.5001
16/16 [==============================] - 0s 927us/step - loss: 2.4717
16/16 [==============================] - 0s 2ms/step - loss: 2.4126
16/16 [==============================] - 0s 809us/step - loss: 2.3509
16/16 [==============================] - 0s 2ms/step - loss: 2.3049
16/16 [==============================] - 0s 824us/step - loss: 2.2877
16/16 [==============================] - 0s 2ms/step - loss: 2.2826
16/16 [==============================] - 0s 2ms/step - loss: 2.2814

Testing for epoch 59 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 876us/step - loss: 0.0630
16/16 [==============================] - 0s 2ms/step - loss: 2.1906
16/16 [==============================] - 0s 810us/step - loss: 2.5453
16/16 [==============================] - 0s 785us/step - loss: 2.5202
16/16 [==============================] - 0s 811us/step - loss: 2.4646
16/16 [==============================] - 0s 1ms/step - loss: 2.4054
16/16 [==============================] - 0s 2ms/step - loss: 2.3608
16/16 [==============================] - 0s 2ms/step - loss: 2.3437
16/16 [==============================] - 0s 2ms/step - loss: 2.3383
16/16 [==============================] - 0s 2ms/step - loss: 2.3369

Testing for epoch 59 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0609
16/16 [==============================] - 0s 802us/step - loss: 2.2288
16/16 [==============================] - 0s 813us/step - loss: 2.5893
16/16 [==============================] - 0s 2ms/step - loss: 2.5610
16/16 [==============================] - 0s 2ms/step - loss: 2.5004
16/16 [==============================] - 0s 2ms/step - loss: 2.4371
16/16 [==============================] - 0s 2ms/step - loss: 2.3899
16/16 [==============================] - 0s 857us/step - loss: 2.3722
16/16 [==============================] - 0s 2ms/step - loss: 2.3669
16/16 [==============================] - 0s 2ms/step - loss: 2.3656

Testing for epoch 59 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 824us/step - loss: 0.0613
16/16 [==============================] - 0s 838us/step - loss: 2.2875
16/16 [==============================] - 0s 2ms/step - loss: 2.6576
16/16 [==============================] - 0s 2ms/step - loss: 2.6298
16/16 [==============================] - 0s 2ms/step - loss: 2.5695
16/16 [==============================] - 0s 2ms/step - loss: 2.5058
16/16 [==============================] - 0s 841us/step - loss: 2.4581
16/16 [==============================] - 0s 2ms/step - loss: 2.4401
16/16 [==============================] - 0s 999us/step - loss: 2.4346
16/16 [==============================] - 0s 2ms/step - loss: 2.4333
Epoch 60 of 60

Testing for epoch 60 index 1:
79/79 [==============================] - 0s 2ms/step
16/16 [==============================] - 0s 865us/step - loss: 0.0597
16/16 [==============================] - 0s 845us/step - loss: 2.2392
16/16 [==============================] - 0s 2ms/step - loss: 2.5978
16/16 [==============================] - 0s 2ms/step - loss: 2.5659
16/16 [==============================] - 0s 1ms/step - loss: 2.5047
16/16 [==============================] - 0s 2ms/step - loss: 2.4410
16/16 [==============================] - 0s 2ms/step - loss: 2.3934
16/16 [==============================] - 0s 1ms/step - loss: 2.3755
16/16 [==============================] - 0s 849us/step - loss: 2.3702
16/16 [==============================] - 0s 824us/step - loss: 2.3690

Testing for epoch 60 index 2:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0642
16/16 [==============================] - 0s 855us/step - loss: 2.1887
16/16 [==============================] - 0s 827us/step - loss: 2.5416
16/16 [==============================] - 0s 2ms/step - loss: 2.5131
16/16 [==============================] - 0s 2ms/step - loss: 2.4568
16/16 [==============================] - 0s 1ms/step - loss: 2.3981
16/16 [==============================] - 0s 2ms/step - loss: 2.3537
16/16 [==============================] - 0s 1ms/step - loss: 2.3373
16/16 [==============================] - 0s 2ms/step - loss: 2.3327
16/16 [==============================] - 0s 2ms/step - loss: 2.3317

Testing for epoch 60 index 3:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0631
16/16 [==============================] - 0s 878us/step - loss: 2.2005
16/16 [==============================] - 0s 825us/step - loss: 2.5564
16/16 [==============================] - 0s 2ms/step - loss: 2.5270
16/16 [==============================] - 0s 905us/step - loss: 2.4684
16/16 [==============================] - 0s 804us/step - loss: 2.4070
16/16 [==============================] - 0s 2ms/step - loss: 2.3608
16/16 [==============================] - 0s 845us/step - loss: 2.3434
16/16 [==============================] - 0s 804us/step - loss: 2.3382
16/16 [==============================] - 0s 812us/step - loss: 2.3369

Testing for epoch 60 index 4:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 2ms/step - loss: 0.0611
16/16 [==============================] - 0s 854us/step - loss: 2.1832
16/16 [==============================] - 0s 2ms/step - loss: 2.5365
16/16 [==============================] - 0s 1ms/step - loss: 2.5057
16/16 [==============================] - 0s 2ms/step - loss: 2.4469
16/16 [==============================] - 0s 1ms/step - loss: 2.3865
16/16 [==============================] - 0s 847us/step - loss: 2.3412
16/16 [==============================] - 0s 2ms/step - loss: 2.3242
16/16 [==============================] - 0s 2ms/step - loss: 2.3192
16/16 [==============================] - 0s 856us/step - loss: 2.3180

Testing for epoch 60 index 5:
79/79 [==============================] - 0s 1ms/step
16/16 [==============================] - 0s 1ms/step - loss: 0.0615
16/16 [==============================] - 0s 840us/step - loss: 2.1932
16/16 [==============================] - 0s 2ms/step - loss: 2.5424
16/16 [==============================] - 0s 1ms/step - loss: 2.5080
16/16 [==============================] - 0s 803us/step - loss: 2.4476
16/16 [==============================] - 0s 2ms/step - loss: 2.3862
16/16 [==============================] - 0s 1ms/step - loss: 2.3406
16/16 [==============================] - 0s 1ms/step - loss: 2.3239
16/16 [==============================] - 0s 1ms/step - loss: 2.3190
16/16 [==============================] - 0s 2ms/step - loss: 2.3179
79/79 [==============================] - 0s 1ms/step

MO_GAAL(contamination=0.05, k=10, lr_d=0.01, lr_g=0.0001, momentum=0.9,
    stop_epochs=20)

outlier_MO_GAAL_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_bunny,outlier_MO_GAAL_one)
_conf.conf("MO-GAAL (Liu et al., 2019)")

Accuracy: 0.952
Precision: 0.000
Recall: 0.000
F1 Score: 0.000

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/sklearn/metrics/_classification.py:1469: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.
  _warn_prf(average, modifier, msg_start, len(result))

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 2.383
Precision(TP / TP + FP): nan
Recall(TP / TP + FN): 0.000
F1 Score(2*precision*recall/precision+recall): nan

/tmp/ipykernel_3852735/4166638268.py:3: RuntimeWarning: invalid value encountered in long_scalars
  print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
/tmp/ipykernel_3852735/4166638268.py:5: RuntimeWarning: invalid value encountered in long_scalars
  print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

fpr, tpr, thresh = roc_curve(outlier_true_bunny,clf.decision_function(_df[['x', 'y','fnoise']]))

79/79 [==============================] - 0s 1ms/step

auc(fpr, tpr)

0.6595030773534759

tab_bunny = pd.concat([tab_bunny,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_bunny

	Accuracy	Precision	Recall	F1	AUC
GODE	0.988414	0.864000	0.900000	0.881633	0.996227
LOF (Breunig et al., 2000)	0.943268	0.412698	0.433333	0.422764	0.819111
kNN (Ramaswamy et al., 2000)	0.987215	0.849206	0.891667	0.869919	0.984438
CBLOF (He et al., 2003)	0.980823	0.785714	0.825000	0.804878	0.970545
OCSVM (Sch ̈olkopf et al., 2001)	0.916500	0.322709	0.675000	0.436658	0.857550
MCD (Hardin and Rocke, 2004)	0.978426	0.761905	0.800000	0.780488	0.972398
Feature Bagging (Lazarevic and Kumar, 2005)	0.948861	0.468254	0.491667	0.479675	0.833809
ABOD (Kriegel et al., 2008)	0.979225	0.769841	0.808333	0.788618	0.971576
Isolation Forest (Liu et al., 2008)	0.972034	0.698413	0.733333	0.715447	0.967954
HBOS (Goldstein and Dengel, 2012)	0.932082	0.301587	0.316667	0.308943	0.859192
SOS (Janssens et al., 2012)	0.908909	0.071429	0.075000	0.073171	0.557431
SO-GAAL (Liu et al., 2019)	0.952058	0.000000	0.000000	0.000000	0.673010
MO-GAAL (Liu et al., 2019)	0.952058	0.000000	0.000000	0.000000	0.659503

LSCP_Bunny

detectors = [KNN(), LOF(), OCSVM()]
clf = LSCP(detectors,contamination=0.05, random_state=77)
clf.fit(_df[['x', 'y','fnoise']])

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/pyod/models/lscp.py:382: UserWarning: The number of histogram bins is greater than the number of classifiers, reducing n_bins to n_clf.
  warnings.warn(

LSCP(contamination=0.05,
   detector_list=[KNN(algorithm='auto', contamination=0.1, leaf_size=30, method='largest',
  metric='minkowski', metric_params=None, n_jobs=1, n_neighbors=5, p=2,
  radius=1.0), LOF(algorithm='auto', contamination=0.1, leaf_size=30, metric='minkowski',
  metric_params=None, n_jobs=1, n_neighbors=20, novelty=True, p=2), OCSVM(cache_size=200, coef0=0.0, contamination=0.1, degree=3, gamma='auto',
   kernel='rbf', max_iter=-1, nu=0.5, shrinking=True, tol=0.001,
   verbose=False)],
   local_max_features=1.0, local_region_size=30, n_bins=3,
   random_state=RandomState(MT19937) at 0x7FD5E1460D40)

outlier_LSCP_one = list(clf.labels_)

_conf = Conf_matrx(outlier_true_bunny,outlier_LSCP_one)
_conf.conf("LSCP (Zhao et al., 2019)")

Accuracy: 0.982
Precision: 0.802
Recall: 0.842
F1 Score: 0.821

# check
print('Accuracy(TP + TN / TP + TN + FP + FN): %.3f' % round((_conf.conf_matrix[0][0] + _conf.conf_matrix[1][1])/1000,3))
print('Precision(TP / TP + FP): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]),3))
print('Recall(TP / TP + FN): %.3f' % round(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]),3))
print('F1 Score(2*precision*recall/precision+recall): %.3f' % round((2*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]))*(_conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1]))) / (_conf.conf_matrix[1][1]/(_conf.conf_matrix[0][1] + _conf.conf_matrix[1][1]) + _conf.conf_matrix[1][1]/(_conf.conf_matrix[1][0] + _conf.conf_matrix[1][1])),3))

Accuracy(TP + TN / TP + TN + FP + FN): 2.459
Precision(TP / TP + FP): 0.802
Recall(TP / TP + FN): 0.842
F1 Score(2*precision*recall/precision+recall): 0.821

fpr, tpr, thresh = roc_curve(outlier_true_bunny,clf.decision_function(_df[['x', 'y','fnoise']]))

/home/csy/anaconda3/envs/pygsp/lib/python3.10/site-packages/scipy/stats/_stats_py.py:4068: PearsonRConstantInputWarning: An input array is constant; the correlation coefficient is not defined.
  warnings.warn(PearsonRConstantInputWarning())

auc(fpr, tpr)

0.982259756609316

tab_bunny = pd.concat([tab_bunny,
           pd.DataFrame({"Accuracy":[_conf.acc],"Precision":[_conf.pre],"Recall":[_conf.rec],"F1":[_conf.f1],"AUC":[auc(fpr, tpr)]},index = [_conf.name])]);tab_bunny

	Accuracy	Precision	Recall	F1	AUC
GODE	0.988414	0.864000	0.900000	0.881633	0.996227
LOF (Breunig et al., 2000)	0.943268	0.412698	0.433333	0.422764	0.819111
kNN (Ramaswamy et al., 2000)	0.987215	0.849206	0.891667	0.869919	0.984438
CBLOF (He et al., 2003)	0.980823	0.785714	0.825000	0.804878	0.970545
OCSVM (Sch ̈olkopf et al., 2001)	0.916500	0.322709	0.675000	0.436658	0.857550
MCD (Hardin and Rocke, 2004)	0.978426	0.761905	0.800000	0.780488	0.972398
Feature Bagging (Lazarevic and Kumar, 2005)	0.948861	0.468254	0.491667	0.479675	0.833809
ABOD (Kriegel et al., 2008)	0.979225	0.769841	0.808333	0.788618	0.971576
Isolation Forest (Liu et al., 2008)	0.972034	0.698413	0.733333	0.715447	0.967954
HBOS (Goldstein and Dengel, 2012)	0.932082	0.301587	0.316667	0.308943	0.859192
SOS (Janssens et al., 2012)	0.908909	0.071429	0.075000	0.073171	0.557431
SO-GAAL (Liu et al., 2019)	0.952058	0.000000	0.000000	0.000000	0.673010
MO-GAAL (Liu et al., 2019)	0.952058	0.000000	0.000000	0.000000	0.659503
LSCP (Zhao et al., 2019)	0.982421	0.801587	0.841667	0.821138	0.982260

tab_bunny

round(tab_bunny,3)

	Accuracy	Precision	Recall	F1	AUC
GODE	0.988	0.864	0.900	0.882	0.996
LOF (Breunig et al., 2000)	0.943	0.413	0.433	0.423	0.819
kNN (Ramaswamy et al., 2000)	0.987	0.849	0.892	0.870	0.984
CBLOF (He et al., 2003)	0.981	0.786	0.825	0.805	0.971
OCSVM (Sch ̈olkopf et al., 2001)	0.917	0.323	0.675	0.437	0.858
MCD (Hardin and Rocke, 2004)	0.978	0.762	0.800	0.780	0.972
Feature Bagging (Lazarevic and Kumar, 2005)	0.949	0.468	0.492	0.480	0.834
ABOD (Kriegel et al., 2008)	0.979	0.770	0.808	0.789	0.972
Isolation Forest (Liu et al., 2008)	0.972	0.698	0.733	0.715	0.968
HBOS (Goldstein and Dengel, 2012)	0.932	0.302	0.317	0.309	0.859
SOS (Janssens et al., 2012)	0.909	0.071	0.075	0.073	0.557
SO-GAAL (Liu et al., 2019)	0.952	0.000	0.000	0.000	0.673
MO-GAAL (Liu et al., 2019)	0.952	0.000	0.000	0.000	0.660
LSCP (Zhao et al., 2019)	0.982	0.802	0.842	0.821	0.982