티스토리 뷰
www.edwith.org/linearalgebra4ai/lecture/23755/
[LECTURE] 선형독립과 선형종속 : edwith
학습목표 이번 강의에서는 선형대수에서 중요한 개념 중 하나인 선형독립과 선형종속에 대해서 배우겠습니다. 그리고 이들이 선형 시스템 내에서 가지는 특성에 대해서도 알아보겠습니다. ...
www.edwith.org
Uniqueness of Solution for \(A\mathbf{x}= \mathbf{b}\)
The solution exists only when \(\mathbf{b} \in \) Span \( \{\mathbf{a}_{1}, \mathbf{a}_{2}, \mathbf{a}_{3} \} \).
$$\begin{bmatrix} 60 \\ 65 \\ 55 \end{bmatrix}x_{1}+\begin{bmatrix} 5.5 \\ 5.0 \\ 6.0 \end{bmatrix}x_{2}+\begin{bmatrix} 1 \\ 0 \\ 1 \end{bmatrix}x_{3} = \begin{bmatrix} 66 \\ 74 \\ 78 \end{bmatrix}$$
$$\mathbf{a}_{1}x_{1}+\mathbf{a}_{2}x_{2}+\mathbf{a}_{3}x_{3}=\mathbf{b}$$
\(\mathbf{b} \in \) Span \( \{\mathbf{a}_{1}, \mathbf{a}_{2}, \mathbf{a}_{3} \} \)일때, Linear System의 Solution 유무를 알 수 있습니다. 그렇지만, 해당 정보만으로는 Solution이 Unique한지, Not Unique한지는 알 수 없습니다.
Linear Independence를 알고나면 위 문제에 대해 대답을 할 수 있습니다.
Linear Independence
(Practical) Definition
Given a set of vectors \(\mathbf{v_{1}}, \cdots, \mathbf{v_{p}} \in \mathbb{R}^{n} \), check if \(\mathbf{v_{j}}\) can be represented as a linear combination of the previous vectors \(\{\mathbf{v_{1}}, \mathbf{v_{2}}, \ldots, \mathbf{v_{j-1}} \}\) for \(j = 1,\ldots, p\), e.g.,
$$ \mathbf{v_{j}} \in \mathsf{Span} \{\mathbf{v_{1}}, \mathbf{v_{2}}, \ldots, \mathbf{v_{j-1}} \} \mathsf{\ for \ some} \ j = 1, \ldots , p? $$
If at least one such \(\mathbf{v_{j}}\) is found, then \(\{\mathbf{v_{1}}, \cdots, \mathbf{v_{p}}\}\) is linearly dependent.
If no such \(\mathbf{v_{j}}\) is found, then \(\{\mathbf{v_{1}}, \cdots, \mathbf{v_{p}}\}\) is linearly independent.
주어진 벡터들의 Set의 어느 한 벡터가 Span{나머지 벡터들}에 존재한다면 해당 벡터들의 Set은 linearly dependent하고, 그러한 벡터가 존재하지 않는다면 해당 벡터들의 Set은 linearly independent하다고 정의할 수 있습니다.
Span이 이루는 벡터의 차원이 n이고 Span을 구성하는 Set of Vectors의 Vector의 갯수가 n보다 큰 경우, 해당 Set of Vectors는 무조건 linearly dependent하다고 정의할 수 있습니다. 이는 Linear System에서 미지수의 갯수보다 방정식의 갯수가 적은 경우를 의미합니다. 이 경우, 무조건 해가 무수히 많이 존재하게 됩니다. 이 경우는 Set of vectors 가 linearly dependent한 경우와 같습니다.
Span이 이루는 벡터의 차원이 n이고 Span을 구성하는 Set of Vectors의 Vector의 갯수가 n보다 작은 경우, 해당 Set of Vectors는 linearly dependent 할 수도 있고, linearly independent 할 수도 있습니다.
(Formal) Definition
Consider \(x_{1}\mathbf{v}_{1} + x_{2}\mathbf{v}_{2} + \cdots + x_{p}\mathbf{v}_{p} = \mathbf{b} = \mathbf{0}.\)
Obviously, one solution is \(\mathbf{x}= \begin{bmatrix} x_{1} \\ x_{2} \\ \vdots \\ x_{p} \end{bmatrix}=\begin{bmatrix} 0 \\ 0 \\ \vdots \\ 0 \end{bmatrix}\),
which we call a trivial(자명한, 하찮은) solution.
영벡터는 모든 Set of Vectors의 Span에 무조건 존재합니다.
\( \mathbf{v}_{1}, \cdots, \mathbf{v}_{p}\) are linearly independent if this(=trivial solution) is the only solution.
\( \mathbf{v}_{1}, \cdots, \mathbf{v}_{p}\) are linearly dependent if this system also has other non trivial solutions, e.g., at least one \(x_{i}\) being nonzero.
Tivial Solution이 아닌 다른 Solution이 존재할때 식\(x_{1}\mathbf{v}_{1} + x_{2}\mathbf{v}_{2} + \cdots + x_{p}\mathbf{v}_{p} = \mathbf{b} = \mathbf{0}.\)을 전개해보면 어떤 벡터 = 나머지 벡터들의 Linear Combination 형태로 표현되므로 해당 Set of Vectors는 linearly dependent하게 됩니다.
Linear Dependence
A linearly dependent vector does not increase Span!
If \( \mathbf{v}_{3} \in \ \mathsf{Span}\{\mathbf{v}_{1}, \mathbf{v}_{2}\} \),then
$$\mathsf{Span}\{\mathbf{v}_{1}, \mathbf{v}_{2}\} = \mathsf{Span}\{\mathbf{v}_{1}, \mathbf{v}_{2}, \mathbf{v}_{3}\}$$
Why?
Suppose \(\mathbf{v}_{3} = d_{1}\mathbf{v}_{1} + d_{2}\mathbf{v}_{2}\), then the linear combination of \(\mathbf{v}_{1}, \mathbf{v}_{2}, \mathbf{v}_{3}\) can be written as
$$c_{1}\mathbf{v}_{1} + c_{2}\mathbf{v}_{2} + c_{3}\mathbf{v}_{p} = (c_{1} + c_{3}d_{1})\mathbf{v}_{1} + (c_{2} + c_{3}d_{2})\mathbf{v}_{2}$$
which is also a linear combination of \(\mathbf{v}_{1}, \mathbf{v}_{2}\).
Linear Dependence and Linear System Solution
\(x_{1}\mathbf{v}_{1} + x_{2}\mathbf{v}_{2} + \cdots + x_{p}\mathbf{v}_{p} = \mathbf{b} = \mathbf{0}.\)
이때, Trivial Solution외에 다른 Solution이 존재하면(= Linear Dependent하면), \(\mathbf{b} \ne \mathbf{0}\)일 때도 Solution이 Not Unique함이 성립됩니다.
'Math > Linear Algebra' 카테고리의 다른 글
| 인공지능을 위한 선형대수 - CHAPTER 2.5 선형변환 (0) | 2020.10.20 |
|---|---|
| 인공지능을 위한 선형대수 - CHAPTER 2.4 부분공간의 기저와 차원 (0) | 2020.10.18 |
| 인공지능을 위한 선형대수 - CHAPTER 2.2 선형결합 (0) | 2020.10.10 |
| 인공지능을 위한 선형대수 - CHAPTER 2.1 선형방정식과 선형시스템 (0) | 2020.10.09 |
| 인공지능을 위한 선형대수 - CHAPTER 1. 선형대수의 기초 (0) | 2020.10.04 |
- Total
- Today
- Yesterday
- 백준 11437
- 인공지능을 위한 선형대수
- PyCharm
- 순열
- cosine
- FairMOT
- C++ Deploy
- 백준 11053
- 이분탐색
- 백준 1766
- ㅂ
- Lowest Common Ancestor
- 백준
- 자료구조
- 백트래킹
- LCA
- 파이참
- MOT
- 위상 정렬 알고리즘
- 조합
- 문제집
- 가장 긴 증가하는 부분 수열
- 단축키
| 일 | 월 | 화 | 수 | 목 | 금 | 토 |
|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | 6 | |
| 7 | 8 | 9 | 10 | 11 | 12 | 13 |
| 14 | 15 | 16 | 17 | 18 | 19 | 20 |
| 21 | 22 | 23 | 24 | 25 | 26 | 27 |
| 28 | 29 | 30 | 31 |
