[Linear Algebra] Lecture 11

Author

SEOYEON CHOI

Published

March 20, 2024

학습 목표

- Bases of all vector spaces

\(3 \times 3\) 행렬의 부분 공간 1. \(3 \times 3\) symmetric matrix 대칭 행렬 2. \(3 \times 3\) upper triangular matrix 상삼각행렬 3. \(3 \times 3\) diagonal matrix 대각 행렬

- basis of \(M\)

$M = $ all 3 bt 3 matrices, dim(M) = 9

Symmetric \(3 \times 3\), dim(S) = 6(diagonal = 3, 3 entries)

matrices upper triangular \(3 \times 3\) dim(U)=6

Basis for \(M=\) all \(3 \times 3\)’s

\(\to\) standard basis1 for \(M\), 9개

  • 1 1과 0으로만 구성된 basis

  • \(\begin{bmatrix} 1 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 1 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 0 & 1 \\ 0 & 0 & 0 \\ 0 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 0 & 0 \\ 1 & 0 & 0 \\ 0 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 0 & 0 \\ 0 & 1 & 0 \\ 0 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 0 & 0 \\ 0 & 0 & 1 \\ 0 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 0 & 0 \\ 0 & 0 & 0 \\ 1 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 1 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 1 \end{bmatrix}\)

    - symmetric matrix as a subspace of \(M\)

    \(S\), symmetrix matrix의 basis에 포함되는 \(M\)의 기저 행렬들, 3개

    \(\begin{bmatrix} 1 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 0 & 0 \\ 0 & 1 & 0 \\ 0 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 1 \end{bmatrix}\)

    symmetrix matrix의 basis인 나머지 3개

    \(\begin{bmatrix} 0 & 1 & 0 \\ 1 & 0 & 0 \\ 0 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 0 & 1 \\ 0 & 0 & 0 \\ 1 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 0 & 0 \\ 0 & 0 & 1 \\ 0 & 1 & 0 \end{bmatrix}\)

    즉, symmetric matrix의 dimension은 6

    -upper triangular matrix as a subspace of \(M\)

    \(U\), Upper truangular matrix, 상삼각행렬의 basis, 기저는 6개

    \(\begin{bmatrix} 1 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 1 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 0 & 1 \\ 0 & 0 & 0 \\ 0 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 0 & 0 \\ 0 & 1 & 0 \\ 0 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 0 & 0 \\ 0 & 0 & 1 \\ 0 & 0 & 0 \end{bmatrix}\), \(\begin{bmatrix} 0 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 1 \end{bmatrix}\)

    즉, 차원은 6

    - diagonal matrix as a subspace of \(M\)

    \(S \cap U\) = symmetric and upper triangular

    = diagonal \(3 \times 3\)’s

    = dim(\(A \cap U\)) = 3

    \(S \cap U = D \to\) \(\begin{bmatrix} 1 & 1& 1 \\1 & 1& 1 \\1 & 1& 1\end{bmatrix}\) $$ \(\begin{bmatrix}1 & 1& 1 \\ 0 & 1& 1 \\ 0 & 0& 1 \\\end{bmatrix}\)=\(\begin{bmatrix} 1 & 0&0 \\0 & 1& 0 \\ 0 & 0& 1 \end{bmatrix}\)

    즉, 대각행렬의 차원이 3이다.

    \(S \cup U\) \(S,U\) have different direction, respectively.

    \(S + U\) = any element of \(S\) + any element of \(U\) = all \(3 \times 3\)’s

    dim(\(S+U\)) = 9

    dim S = 6 + dim \(U\) = 6

    sum of dimension = 6 + 6 = 3 + 9 (\(\cap + \cup\))

    -> 따라서 부분공간으로 정의하는 차원이나 기저 등의 연산이 자연스럽게 산출된다는 증거로 볼 수 있다.

    Example

    미분 방정식을 예로 들자면,

    \(\frac{d^2 y}{d x^2} + y = 0\)

    이 basis는

    \(\frac{d^2 y}{d x^2} + y = y^{''} + y = 0\)

    이렇게 보니, 풀어쓴 해는 basis의 영공간nullspace를 나타낸 것으로도 볼 수 있다.

    complete solution, 즉 완전해를 정의하자면?

  • 2 all combination 표현할 수 있으면 basis로 보기로 했잖아

  • 즉, 미분방정식은 \(Ax=0\)의 solution으로 해석 가능하고, 선형 조합한 결과(3)는 영공간nullspace를 정의하기 위한 special solution으로 볼 수 있다.

    dim(solution space) = 2

    - Rank one matrices

    Suppose rank = 1, this is the exmaple.

    \(A = \begin{bmatrix} 1 & 4 & 5 \\ 2 & 8 & 10 \end{bmatrix}\), 2 by 3 matrix

    row 2가 row 1의 두 배가 되어 서로 종속인 상태가 되기 때문에 rank 1인 예시로 가능하다. col 2도 col 1에 4배를 한 것과 같고, col 3도 col 1에 5배를 한 것과 같아 종속 상태에 있다.

    \(\begin{bmatrix} 1 \\ 2 \end{bmatrix}\) \(\times\) \(\begin{bmatrix} 1 & 4 & 5\end{bmatrix}\) = \(\begin{bmatrix} 1 & 4 & 5 \\ 2 & 8 & 10 \end{bmatrix}\) = \(A\)

    \(C(A) \times C(A)^T = A\)

    \(\therefore\) Rank 1 matrix: \(A = uv^T\)

    column vector \(\times\) row vector 를 곱하면 꼭 rank 1 행렬 A 가 만들어진다.

    -> rank 1 의 조합으로 행렬 표현 가능

    - Subspaces for rank 1 matrices

    \(M =\) all \(5 \times 14\) matrices

    \(\begin{bmatrix} 1 & 4 & 5 \\ 2 & 8 & 10 \end{bmatrix}\) + \(\begin{bmatrix} 4 & 4 & 2 \\ 2 & 1 & 10 \end{bmatrix}\) = \(\begin{bmatrix} 5 & 8 & 7 \\ 4 & 9 & 20 \end{bmatrix}\)

    rank 1 + rank 2 = rank 2

    - in \(R^4\) cases

    \(v = \begin{bmatrix} v_1 \\ v_2 \\ v_3 \\ v_4\end{bmatrix}\)

    \(R^4\)의 부분공간 \(S\) = all \(v\) in \(R^4\) with \(v_1 + v_2 + v_3 + v_4 = 0\)을 정의해보면 이렇다.

    \(A = \begin{bmatrix} 1 & 1 & 1 & 1\end{bmatrix}\)

    \(Av=0\)

    special soultions of null space

    \(\begin{bmatrix} -1 \\ 1 \\ 0 \\ 0\end{bmatrix}\), \(\begin{bmatrix} -1 \\ 0 \\ 1 \\ 0\end{bmatrix}\), \(\begin{bmatrix} -1 \\ 0 \\ 0 \\ 1\end{bmatrix}\)

    \(Ax = 0\), \(\begin{bmatrix} 1\\1\\1\\1 \end{bmatrix}\) \(\begin{bmatrix} x \end{bmatrix}\) = \(\begin{bmatrix} 0\\0\\0\\0 \end{bmatrix}\)

    row space(1) + left null space(0) \(\times\) column space(1) + null space(3) = 1 \(\times\) 4 \(\to\) A 행렬의 차원이 됌