SlideShare a Scribd company logo
Linear Algebra
9. Linear Transformation
한양대 이상화 교수님 <선형대수>
http://www.kocw.net/home/search/kemView.do?kemId=977757
Linear Transformation
• 이제까지 𝐴𝑥 = 𝑏는 ‘어떤 Input 𝒙를 넣었을 때 Output 𝒃가 나오는 시스템 𝑨’ 로 이해하였다.
(𝐴의 열벡터들을 각자 다른 𝑥로 조합을 하면 특정한 𝑏가 나오는 형태 : 𝑏 is a linear combination of
column vectors of 𝐴 with coefficients in 𝑥)
• 𝐴𝑥 = 𝑏 (𝐴는 𝑚 × 𝑛 행렬)를 새로운 관점에서 해석해보자.
‘𝑨𝒙 = 𝒃 는 n차원의 Input 𝒙를 m차원의 Output 𝒃로 변환하는 과정’
( 𝑥 is transformed/mapped into 𝑏 by 𝐴)
AInput 𝒙 Output 𝒃
𝒙 ∈ 𝑹 𝒏 𝒃 ∈ 𝑹 𝒎
𝑨
• 예시
𝐴 =
𝑐 0
0 𝑐
 Stretching (Extending or Contracting)
𝐴 =
0 −1
1 0
 90° Rotation
𝐴 =
0 1
1 0
 Reflection by 𝑥2 = 𝑥1
𝐴 =
1 0
0 0
 Projection onto x
Linear Transformation
• 𝑻 𝒙 = 𝑨𝒙
• 원점은 이동할 수 없다. 모든 𝐴에 대해 𝐴𝑥 = 𝐴 ∙ 0 = 0
• 𝐴(𝑐𝑥) = 𝑐(𝐴𝑥)
• 𝐴 𝑥 + 𝑦 = 𝐴𝑥 + 𝐴𝑦
 𝑨(𝒂𝒙 + 𝒃𝒚) = 𝒂(𝑨𝒙) + 𝒃(𝑨𝒚)
Linear Transformation
A matrix Linear Transformation
O
O
행렬은 선형변환 과정으로
이해할 수 있고,
모든 선형변환은 행렬로
표현할 수 있다.
• 선형 변환에서 x가 반드시 우리가 생각하는 단순한 벡터일 필요는 없다.
• 𝑥가 함수일 경우, 𝑃 𝑡 = 𝑎0 + 𝑎1 𝑡 + 𝑎2 𝑡2
+ ⋯ + 𝑎 𝑛 𝑡 𝑛
 rank = n+1
• Differentiation : 𝐴 = 𝑑/𝑑𝑡 is Linear
𝐴𝑃 𝑡 = 𝑎0 + 2𝑎1 𝑡 + ⋯ + 𝑛𝑎 𝑛 𝑡 𝑛−1
 rank = n
• Integration
𝐴𝑃 𝑡 = 0
𝑡
𝑎0 + 𝑎1 𝑡 + ⋯ 𝑎 𝑛 𝑡 𝑛
𝑑𝑡 = 𝑎0 𝑡 + ⋯ +
𝑎 𝑛
𝑛+1
𝑡 𝑛+1
선형 함수는 모두 행렬로 표현할 수 있다.
• 모든 Basis 벡터에 대해 𝑨𝒙 값을 알고 있으면 벡터 공간 내의 모든 𝒙에 대해 𝑨𝒙 값을 알 수 있다.
• 벡터 공간의 모든 𝒙는 Basis에 의해 표현될 수 있기 때문에 𝑨를 몰라도 𝑨𝒙를 구할 수 있다.
𝒙 = 𝒄 𝟏 𝒙 𝟏 + 𝒄 𝟐 𝒙 𝟐 + ⋯ + 𝒄 𝒏 𝒙 𝒏
𝑨𝒙 = 𝑨 𝒄 𝟏 𝒙 𝟏 + 𝒄 𝟐 𝒙 𝟐 + ⋯ + 𝒄 𝒏 𝒙 𝒏 = 𝒄 𝟏 𝑨𝒙 𝟏 + 𝒄 𝟐 𝑨𝒙 𝟐 + ⋯ + 𝒄 𝒏 𝑨𝒙 𝒏
Linear Transformation
다항식도 벡터로 이해할 수 있다.
(계수들의 N+1 차원 벡터)
• Elementary Basis Vector들의 Ax 값을 안다면 A는 쉽게 구할 수 있다.
𝑥1 =
1
0
, 𝐴𝑥1 =
2
3
4
𝑥2 =
0
1
, 𝐴𝑥1 =
4
6
8
Elementary Basis가 아닌 (1, 1), (2, -1)에 대해 같은 방법으로 A를 다시 구해보면,
A
1
1
= 𝐴 𝑥1 + 𝑥2 =
4
6
8
+
2
3
4
=
6
9
12
A
2
−1
= 𝐴 2𝑥1 − 𝑥2 =
4
6
8
−
4
6
8
=
0
0
0
Linear Transformation Matrix A
𝑇 𝑥 = 𝐴 =
2 4
3 6
4 8
, 𝐴
−1
−2
=
−10
−15
−20
𝐴′ =
6 0
9 0
12 0
, 𝐴′
−1
−2
=
−6
−9
−12
값이 다름
위 방법은
Elementary
에만 적용할
수 있다.
(-1, -2)를 basis
1
1
와
2
−1
로 표현해야
• Polynomial Case : Differentiation
𝑥 𝑡 = 𝑎0 + 𝑎1 𝑡 + 𝑎2 𝑡2
+ 𝑎3 𝑡3
일 때,
𝑡0
𝑡1
𝑡2
𝑡3
미지수 위치에 1을 넣은 Elementary Basis를 구성한다.
(𝑃1 = 1, 𝑃2 = 𝑡, 𝑃3 = 𝑡2
, 𝑃4 = 𝑡3
)
미분함수 𝐴 = 𝑑/𝑑𝑡  𝐴𝑃1 = 0, 𝐴𝑃2 = 1 = 𝑃1, 𝐴𝑃3 = 2𝑡 = 2𝑃2, 𝐴𝑃4 = 3𝑡3
= 3𝑃3
𝐴𝑃1 = 𝐴
1
0
0
0

0
0
0
0
, 𝐴𝑃2 = 𝐴
0
1
0
0

1
0
0
0
, 𝐴𝑃3 = 𝐴
0
0
1
0

0
2
0
0
, 𝐴𝑃4 = 𝐴
0
0
0
1

0
0
3
0
𝑃 𝑡 = 2 + 𝑡 − 𝑡2
− 𝑡3
일 때
𝐴𝑃 =
0 1 0 0
0 0 2 0
0 0 0 3
0 0 0 0
2
1
−1
−1
=
1
−2
−3
0
 𝑃′ 𝑡 = 1 − 2𝑡 − 3𝑡2
Linear Transformation Matrix A
𝐴 =
0 1 0 0
0 0 2 0
0 0 0 3
0 0 0 0
𝑃1
𝑃2
𝑃3
𝑃4
• Polynomial Case : Integration
𝑥 𝑡 = 𝑎0 + 𝑎1 𝑡 + 𝑎2 𝑡2
+ 𝑎3 𝑡3
일 때,
적분 함수 𝐴 = 0
𝑡
𝑑𝑡  basis ∶ {1, 𝑡, 𝑡2, 𝑡3}
𝐴𝑃1 = 𝑡 = 𝑃2, 𝐴𝑃2 = 1
2 𝑡2
= 1
2 𝑃3, 𝐴𝑃3 = 1
3 𝑡3
= 1
3 𝑃4, 𝐴𝑃4 = 1
4 𝑡4
= 1
4 𝑃5
𝐴𝑃1 = 𝐴
1
0
0
0

0
1
0
0
0
, 𝐴𝑃2 = 𝐴
0
1
0
0

0
0
1
2
0
0
, 𝐴𝑃3 = 𝐴
0
0
1
0

0
0
0
1
3
0
, 𝐴𝑃4 = 𝐴
0
0
0
1

0
0
0
0
1
4
Linear Transformation Matrix A
𝐴 =
0 0 0 0
1 0 0 0
0 1
2 0 0
0 0 1
3 0
0 0 0 1
4
𝐴 𝑑𝑖𝑓𝑓 𝐴𝑖𝑛𝑡 =
0 1 0 0 0
0 0 2 0 0
0 0 0 3 0
0 0 0 0 4
0 0 0 0
1 0 0 0
0 1
2 0 0
0 0 1
3 0
0 0 0 1
4
=
1 0 0 0
0 1 0 0
0 0 1 0
0 0 0 1
* 적분했다가 미분하면 같은 값
* 𝑨𝒊𝒏𝒕 𝑨 𝒅𝒊𝒇𝒇 ≠ 𝑰
 미분했다가 적분하면 상수항이
생겨 같은 값 아님
* Differentiation is a left inverse
of Integration.

More Related Content

What's hot

Eigen values and eigen vectors engineering
Eigen values and eigen vectors engineeringEigen values and eigen vectors engineering
Eigen values and eigen vectors engineering
shubham211
 
system linear equations and matrices
 system linear equations and matrices system linear equations and matrices
system linear equations and matrices
Aditya Vaishampayan
 
Power Series - Legendre Polynomial - Bessel's Equation
Power Series - Legendre Polynomial - Bessel's EquationPower Series - Legendre Polynomial - Bessel's Equation
Power Series - Legendre Polynomial - Bessel's Equation
ArijitDhali
 
선형대수 06. 영벡터공간과 해집합
선형대수 06. 영벡터공간과 해집합선형대수 06. 영벡터공간과 해집합
선형대수 06. 영벡터공간과 해집합
AHRA CHO
 
03 Machine Learning Linear Algebra
03 Machine Learning Linear Algebra03 Machine Learning Linear Algebra
03 Machine Learning Linear Algebra
Andres Mendez-Vazquez
 
선형대수 01. 선형성의 정의와 1차 연립방정식의 의미
선형대수 01. 선형성의 정의와 1차 연립방정식의 의미선형대수 01. 선형성의 정의와 1차 연립방정식의 의미
선형대수 01. 선형성의 정의와 1차 연립방정식의 의미
AHRA CHO
 
Relaxation method
Relaxation methodRelaxation method
Relaxation method
Parinda Rajapaksha
 
Euler lagrange equations of motion mit-holonomic constraints_lecture7
Euler lagrange equations of motion  mit-holonomic  constraints_lecture7Euler lagrange equations of motion  mit-holonomic  constraints_lecture7
Euler lagrange equations of motion mit-holonomic constraints_lecture7
JOHN OBIDI
 
Complex function
Complex functionComplex function
Complex function
Shrey Patel
 
Gaussian elimination method & homogeneous linear equation
Gaussian elimination method & homogeneous linear equationGaussian elimination method & homogeneous linear equation
Gaussian elimination method & homogeneous linear equation
Student
 
Analytical geometry
Analytical geometryAnalytical geometry
Analytical geometry
KalaivananRaja
 
Matrices
MatricesMatrices
가깝고도 먼 Trpo
가깝고도 먼 Trpo가깝고도 먼 Trpo
가깝고도 먼 Trpo
Woong won Lee
 
Vector space - subspace By Jatin Dhola
Vector space - subspace By Jatin DholaVector space - subspace By Jatin Dhola
Vector space - subspace By Jatin Dhola
Jatin Dhola
 
Lu decomposition
Lu decompositionLu decomposition
Lu decompositiongilandio
 
linear equation and gaussian elimination
linear equation and gaussian eliminationlinear equation and gaussian elimination
linear equation and gaussian elimination
Aju Thadikulangara
 
Gauss Jorden and Gauss Elimination method.pptx
Gauss Jorden and Gauss Elimination method.pptxGauss Jorden and Gauss Elimination method.pptx
Gauss Jorden and Gauss Elimination method.pptx
AHSANMEHBOOB12
 
My Lecture Notes from Linear Algebra
My Lecture Notes fromLinear AlgebraMy Lecture Notes fromLinear Algebra
My Lecture Notes from Linear Algebra
Paul R. Martin
 
Master Thesis on the Mathematial Analysis of Neural Networks
Master Thesis on the Mathematial Analysis of Neural NetworksMaster Thesis on the Mathematial Analysis of Neural Networks
Master Thesis on the Mathematial Analysis of Neural Networks
Alina Leidinger
 
Logistic Regression | Logistic Regression In Python | Machine Learning Algori...
Logistic Regression | Logistic Regression In Python | Machine Learning Algori...Logistic Regression | Logistic Regression In Python | Machine Learning Algori...
Logistic Regression | Logistic Regression In Python | Machine Learning Algori...
Simplilearn
 

What's hot (20)

Eigen values and eigen vectors engineering
Eigen values and eigen vectors engineeringEigen values and eigen vectors engineering
Eigen values and eigen vectors engineering
 
system linear equations and matrices
 system linear equations and matrices system linear equations and matrices
system linear equations and matrices
 
Power Series - Legendre Polynomial - Bessel's Equation
Power Series - Legendre Polynomial - Bessel's EquationPower Series - Legendre Polynomial - Bessel's Equation
Power Series - Legendre Polynomial - Bessel's Equation
 
선형대수 06. 영벡터공간과 해집합
선형대수 06. 영벡터공간과 해집합선형대수 06. 영벡터공간과 해집합
선형대수 06. 영벡터공간과 해집합
 
03 Machine Learning Linear Algebra
03 Machine Learning Linear Algebra03 Machine Learning Linear Algebra
03 Machine Learning Linear Algebra
 
선형대수 01. 선형성의 정의와 1차 연립방정식의 의미
선형대수 01. 선형성의 정의와 1차 연립방정식의 의미선형대수 01. 선형성의 정의와 1차 연립방정식의 의미
선형대수 01. 선형성의 정의와 1차 연립방정식의 의미
 
Relaxation method
Relaxation methodRelaxation method
Relaxation method
 
Euler lagrange equations of motion mit-holonomic constraints_lecture7
Euler lagrange equations of motion  mit-holonomic  constraints_lecture7Euler lagrange equations of motion  mit-holonomic  constraints_lecture7
Euler lagrange equations of motion mit-holonomic constraints_lecture7
 
Complex function
Complex functionComplex function
Complex function
 
Gaussian elimination method & homogeneous linear equation
Gaussian elimination method & homogeneous linear equationGaussian elimination method & homogeneous linear equation
Gaussian elimination method & homogeneous linear equation
 
Analytical geometry
Analytical geometryAnalytical geometry
Analytical geometry
 
Matrices
MatricesMatrices
Matrices
 
가깝고도 먼 Trpo
가깝고도 먼 Trpo가깝고도 먼 Trpo
가깝고도 먼 Trpo
 
Vector space - subspace By Jatin Dhola
Vector space - subspace By Jatin DholaVector space - subspace By Jatin Dhola
Vector space - subspace By Jatin Dhola
 
Lu decomposition
Lu decompositionLu decomposition
Lu decomposition
 
linear equation and gaussian elimination
linear equation and gaussian eliminationlinear equation and gaussian elimination
linear equation and gaussian elimination
 
Gauss Jorden and Gauss Elimination method.pptx
Gauss Jorden and Gauss Elimination method.pptxGauss Jorden and Gauss Elimination method.pptx
Gauss Jorden and Gauss Elimination method.pptx
 
My Lecture Notes from Linear Algebra
My Lecture Notes fromLinear AlgebraMy Lecture Notes fromLinear Algebra
My Lecture Notes from Linear Algebra
 
Master Thesis on the Mathematial Analysis of Neural Networks
Master Thesis on the Mathematial Analysis of Neural NetworksMaster Thesis on the Mathematial Analysis of Neural Networks
Master Thesis on the Mathematial Analysis of Neural Networks
 
Logistic Regression | Logistic Regression In Python | Machine Learning Algori...
Logistic Regression | Logistic Regression In Python | Machine Learning Algori...Logistic Regression | Logistic Regression In Python | Machine Learning Algori...
Logistic Regression | Logistic Regression In Python | Machine Learning Algori...
 

Similar to 선형대수 08. 선형 변환 (Linear Transformation)

Eigendecomposition and pca
Eigendecomposition and pcaEigendecomposition and pca
Eigendecomposition and pca
Jinhwan Suk
 
Neural network (perceptron)
Neural network (perceptron)Neural network (perceptron)
Neural network (perceptron)
Jeonghun Yoon
 
0131 1 spectral_theorem_transformation
0131 1 spectral_theorem_transformation0131 1 spectral_theorem_transformation
0131 1 spectral_theorem_transformation
Jeonghun Yoon
 
0124 1 linear_algebra_basic_vector
0124 1 linear_algebra_basic_vector0124 1 linear_algebra_basic_vector
0124 1 linear_algebra_basic_vector
Jeonghun Yoon
 
RLCode와 A3C 쉽고 깊게 이해하기
RLCode와 A3C 쉽고 깊게 이해하기RLCode와 A3C 쉽고 깊게 이해하기
RLCode와 A3C 쉽고 깊게 이해하기
Woong won Lee
 
Linear algebra.pptx
Linear algebra.pptxLinear algebra.pptx
Linear algebra.pptx
GeonWooYoo1
 
선형대수 04. Inverse and Transpose
선형대수 04. Inverse and Transpose선형대수 04. Inverse and Transpose
선형대수 04. Inverse and Transpose
AHRA CHO
 
04. logistic regression ( 로지스틱 회귀 )
04. logistic regression ( 로지스틱 회귀 )04. logistic regression ( 로지스틱 회귀 )
04. logistic regression ( 로지스틱 회귀 )
Jeonghun Yoon
 
03. linear regression
03. linear regression03. linear regression
03. linear regression
Jeonghun Yoon
 
Deep Learning from scratch 5장 : backpropagation
 Deep Learning from scratch 5장 : backpropagation Deep Learning from scratch 5장 : backpropagation
Deep Learning from scratch 5장 : backpropagation
JinSooKim80
 
Auto-Encoders and Variational Auto-Encoders
Auto-Encoders and Variational Auto-EncodersAuto-Encoders and Variational Auto-Encoders
Auto-Encoders and Variational Auto-Encoders
Jinho Lee
 
Variational AutoEncoder(VAE)
Variational AutoEncoder(VAE)Variational AutoEncoder(VAE)
Variational AutoEncoder(VAE)
강민국 강민국
 
Ray-triangle intersection
Ray-triangle intersectionRay-triangle intersection
Ray-triangle intersection
동환 김
 
06. graph mining
06. graph mining06. graph mining
06. graph mining
Jeonghun Yoon
 
08. spectal clustering
08. spectal clustering08. spectal clustering
08. spectal clustering
Jeonghun Yoon
 
SVM
SVMSVM
0124 2 linear_algebra_basic_matrix
0124 2 linear_algebra_basic_matrix0124 2 linear_algebra_basic_matrix
0124 2 linear_algebra_basic_matrix
Jeonghun Yoon
 

Similar to 선형대수 08. 선형 변환 (Linear Transformation) (20)

Eigendecomposition and pca
Eigendecomposition and pcaEigendecomposition and pca
Eigendecomposition and pca
 
Neural network (perceptron)
Neural network (perceptron)Neural network (perceptron)
Neural network (perceptron)
 
0131 1 spectral_theorem_transformation
0131 1 spectral_theorem_transformation0131 1 spectral_theorem_transformation
0131 1 spectral_theorem_transformation
 
0124 1 linear_algebra_basic_vector
0124 1 linear_algebra_basic_vector0124 1 linear_algebra_basic_vector
0124 1 linear_algebra_basic_vector
 
RLCode와 A3C 쉽고 깊게 이해하기
RLCode와 A3C 쉽고 깊게 이해하기RLCode와 A3C 쉽고 깊게 이해하기
RLCode와 A3C 쉽고 깊게 이해하기
 
Linear algebra.pptx
Linear algebra.pptxLinear algebra.pptx
Linear algebra.pptx
 
선형대수 04. Inverse and Transpose
선형대수 04. Inverse and Transpose선형대수 04. Inverse and Transpose
선형대수 04. Inverse and Transpose
 
04. logistic regression ( 로지스틱 회귀 )
04. logistic regression ( 로지스틱 회귀 )04. logistic regression ( 로지스틱 회귀 )
04. logistic regression ( 로지스틱 회귀 )
 
03. linear regression
03. linear regression03. linear regression
03. linear regression
 
Deep Learning from scratch 5장 : backpropagation
 Deep Learning from scratch 5장 : backpropagation Deep Learning from scratch 5장 : backpropagation
Deep Learning from scratch 5장 : backpropagation
 
Auto-Encoders and Variational Auto-Encoders
Auto-Encoders and Variational Auto-EncodersAuto-Encoders and Variational Auto-Encoders
Auto-Encoders and Variational Auto-Encoders
 
행렬
행렬행렬
행렬
 
Variational AutoEncoder(VAE)
Variational AutoEncoder(VAE)Variational AutoEncoder(VAE)
Variational AutoEncoder(VAE)
 
행렬
행렬행렬
행렬
 
Ray-triangle intersection
Ray-triangle intersectionRay-triangle intersection
Ray-triangle intersection
 
06. graph mining
06. graph mining06. graph mining
06. graph mining
 
08. spectal clustering
08. spectal clustering08. spectal clustering
08. spectal clustering
 
SVM
SVMSVM
SVM
 
벡터
벡터벡터
벡터
 
0124 2 linear_algebra_basic_matrix
0124 2 linear_algebra_basic_matrix0124 2 linear_algebra_basic_matrix
0124 2 linear_algebra_basic_matrix
 

More from AHRA CHO

선형대수 11강 벡터 투영과 최소제곱법
선형대수 11강 벡터 투영과 최소제곱법선형대수 11강 벡터 투영과 최소제곱법
선형대수 11강 벡터 투영과 최소제곱법
AHRA CHO
 
선형대수 07. 선형독립, 4가지 부벡터공간
선형대수 07. 선형독립, 4가지 부벡터공간선형대수 07. 선형독립, 4가지 부벡터공간
선형대수 07. 선형독립, 4가지 부벡터공간
AHRA CHO
 
선형대수 03. LU Decomposition
선형대수 03. LU Decomposition선형대수 03. LU Decomposition
선형대수 03. LU Decomposition
AHRA CHO
 
선형대수 02. 가우스 소거법
선형대수 02. 가우스 소거법선형대수 02. 가우스 소거법
선형대수 02. 가우스 소거법
AHRA CHO
 
01_ML 기초_선형회귀_Linear Regression
01_ML 기초_선형회귀_Linear Regression01_ML 기초_선형회귀_Linear Regression
01_ML 기초_선형회귀_Linear Regression
AHRA CHO
 
00_ML 기초_기본개념
00_ML 기초_기본개념00_ML 기초_기본개념
00_ML 기초_기본개념
AHRA CHO
 
[2012 2] 1주차 사회적경제 개요
[2012 2] 1주차 사회적경제 개요[2012 2] 1주차 사회적경제 개요
[2012 2] 1주차 사회적경제 개요AHRA CHO
 
[2012 2] 1주차 사회적경제의 역사
[2012 2] 1주차 사회적경제의 역사[2012 2] 1주차 사회적경제의 역사
[2012 2] 1주차 사회적경제의 역사
AHRA CHO
 

More from AHRA CHO (8)

선형대수 11강 벡터 투영과 최소제곱법
선형대수 11강 벡터 투영과 최소제곱법선형대수 11강 벡터 투영과 최소제곱법
선형대수 11강 벡터 투영과 최소제곱법
 
선형대수 07. 선형독립, 4가지 부벡터공간
선형대수 07. 선형독립, 4가지 부벡터공간선형대수 07. 선형독립, 4가지 부벡터공간
선형대수 07. 선형독립, 4가지 부벡터공간
 
선형대수 03. LU Decomposition
선형대수 03. LU Decomposition선형대수 03. LU Decomposition
선형대수 03. LU Decomposition
 
선형대수 02. 가우스 소거법
선형대수 02. 가우스 소거법선형대수 02. 가우스 소거법
선형대수 02. 가우스 소거법
 
01_ML 기초_선형회귀_Linear Regression
01_ML 기초_선형회귀_Linear Regression01_ML 기초_선형회귀_Linear Regression
01_ML 기초_선형회귀_Linear Regression
 
00_ML 기초_기본개념
00_ML 기초_기본개념00_ML 기초_기본개념
00_ML 기초_기본개념
 
[2012 2] 1주차 사회적경제 개요
[2012 2] 1주차 사회적경제 개요[2012 2] 1주차 사회적경제 개요
[2012 2] 1주차 사회적경제 개요
 
[2012 2] 1주차 사회적경제의 역사
[2012 2] 1주차 사회적경제의 역사[2012 2] 1주차 사회적경제의 역사
[2012 2] 1주차 사회적경제의 역사
 

선형대수 08. 선형 변환 (Linear Transformation)

  • 1. Linear Algebra 9. Linear Transformation 한양대 이상화 교수님 <선형대수> http://www.kocw.net/home/search/kemView.do?kemId=977757
  • 2. Linear Transformation • 이제까지 𝐴𝑥 = 𝑏는 ‘어떤 Input 𝒙를 넣었을 때 Output 𝒃가 나오는 시스템 𝑨’ 로 이해하였다. (𝐴의 열벡터들을 각자 다른 𝑥로 조합을 하면 특정한 𝑏가 나오는 형태 : 𝑏 is a linear combination of column vectors of 𝐴 with coefficients in 𝑥) • 𝐴𝑥 = 𝑏 (𝐴는 𝑚 × 𝑛 행렬)를 새로운 관점에서 해석해보자. ‘𝑨𝒙 = 𝒃 는 n차원의 Input 𝒙를 m차원의 Output 𝒃로 변환하는 과정’ ( 𝑥 is transformed/mapped into 𝑏 by 𝐴) AInput 𝒙 Output 𝒃 𝒙 ∈ 𝑹 𝒏 𝒃 ∈ 𝑹 𝒎 𝑨
  • 3. • 예시 𝐴 = 𝑐 0 0 𝑐  Stretching (Extending or Contracting) 𝐴 = 0 −1 1 0  90° Rotation 𝐴 = 0 1 1 0  Reflection by 𝑥2 = 𝑥1 𝐴 = 1 0 0 0  Projection onto x Linear Transformation
  • 4. • 𝑻 𝒙 = 𝑨𝒙 • 원점은 이동할 수 없다. 모든 𝐴에 대해 𝐴𝑥 = 𝐴 ∙ 0 = 0 • 𝐴(𝑐𝑥) = 𝑐(𝐴𝑥) • 𝐴 𝑥 + 𝑦 = 𝐴𝑥 + 𝐴𝑦  𝑨(𝒂𝒙 + 𝒃𝒚) = 𝒂(𝑨𝒙) + 𝒃(𝑨𝒚) Linear Transformation A matrix Linear Transformation O O 행렬은 선형변환 과정으로 이해할 수 있고, 모든 선형변환은 행렬로 표현할 수 있다.
  • 5. • 선형 변환에서 x가 반드시 우리가 생각하는 단순한 벡터일 필요는 없다. • 𝑥가 함수일 경우, 𝑃 𝑡 = 𝑎0 + 𝑎1 𝑡 + 𝑎2 𝑡2 + ⋯ + 𝑎 𝑛 𝑡 𝑛  rank = n+1 • Differentiation : 𝐴 = 𝑑/𝑑𝑡 is Linear 𝐴𝑃 𝑡 = 𝑎0 + 2𝑎1 𝑡 + ⋯ + 𝑛𝑎 𝑛 𝑡 𝑛−1  rank = n • Integration 𝐴𝑃 𝑡 = 0 𝑡 𝑎0 + 𝑎1 𝑡 + ⋯ 𝑎 𝑛 𝑡 𝑛 𝑑𝑡 = 𝑎0 𝑡 + ⋯ + 𝑎 𝑛 𝑛+1 𝑡 𝑛+1 선형 함수는 모두 행렬로 표현할 수 있다. • 모든 Basis 벡터에 대해 𝑨𝒙 값을 알고 있으면 벡터 공간 내의 모든 𝒙에 대해 𝑨𝒙 값을 알 수 있다. • 벡터 공간의 모든 𝒙는 Basis에 의해 표현될 수 있기 때문에 𝑨를 몰라도 𝑨𝒙를 구할 수 있다. 𝒙 = 𝒄 𝟏 𝒙 𝟏 + 𝒄 𝟐 𝒙 𝟐 + ⋯ + 𝒄 𝒏 𝒙 𝒏 𝑨𝒙 = 𝑨 𝒄 𝟏 𝒙 𝟏 + 𝒄 𝟐 𝒙 𝟐 + ⋯ + 𝒄 𝒏 𝒙 𝒏 = 𝒄 𝟏 𝑨𝒙 𝟏 + 𝒄 𝟐 𝑨𝒙 𝟐 + ⋯ + 𝒄 𝒏 𝑨𝒙 𝒏 Linear Transformation 다항식도 벡터로 이해할 수 있다. (계수들의 N+1 차원 벡터)
  • 6. • Elementary Basis Vector들의 Ax 값을 안다면 A는 쉽게 구할 수 있다. 𝑥1 = 1 0 , 𝐴𝑥1 = 2 3 4 𝑥2 = 0 1 , 𝐴𝑥1 = 4 6 8 Elementary Basis가 아닌 (1, 1), (2, -1)에 대해 같은 방법으로 A를 다시 구해보면, A 1 1 = 𝐴 𝑥1 + 𝑥2 = 4 6 8 + 2 3 4 = 6 9 12 A 2 −1 = 𝐴 2𝑥1 − 𝑥2 = 4 6 8 − 4 6 8 = 0 0 0 Linear Transformation Matrix A 𝑇 𝑥 = 𝐴 = 2 4 3 6 4 8 , 𝐴 −1 −2 = −10 −15 −20 𝐴′ = 6 0 9 0 12 0 , 𝐴′ −1 −2 = −6 −9 −12 값이 다름 위 방법은 Elementary 에만 적용할 수 있다. (-1, -2)를 basis 1 1 와 2 −1 로 표현해야
  • 7. • Polynomial Case : Differentiation 𝑥 𝑡 = 𝑎0 + 𝑎1 𝑡 + 𝑎2 𝑡2 + 𝑎3 𝑡3 일 때, 𝑡0 𝑡1 𝑡2 𝑡3 미지수 위치에 1을 넣은 Elementary Basis를 구성한다. (𝑃1 = 1, 𝑃2 = 𝑡, 𝑃3 = 𝑡2 , 𝑃4 = 𝑡3 ) 미분함수 𝐴 = 𝑑/𝑑𝑡  𝐴𝑃1 = 0, 𝐴𝑃2 = 1 = 𝑃1, 𝐴𝑃3 = 2𝑡 = 2𝑃2, 𝐴𝑃4 = 3𝑡3 = 3𝑃3 𝐴𝑃1 = 𝐴 1 0 0 0  0 0 0 0 , 𝐴𝑃2 = 𝐴 0 1 0 0  1 0 0 0 , 𝐴𝑃3 = 𝐴 0 0 1 0  0 2 0 0 , 𝐴𝑃4 = 𝐴 0 0 0 1  0 0 3 0 𝑃 𝑡 = 2 + 𝑡 − 𝑡2 − 𝑡3 일 때 𝐴𝑃 = 0 1 0 0 0 0 2 0 0 0 0 3 0 0 0 0 2 1 −1 −1 = 1 −2 −3 0  𝑃′ 𝑡 = 1 − 2𝑡 − 3𝑡2 Linear Transformation Matrix A 𝐴 = 0 1 0 0 0 0 2 0 0 0 0 3 0 0 0 0 𝑃1 𝑃2 𝑃3 𝑃4
  • 8. • Polynomial Case : Integration 𝑥 𝑡 = 𝑎0 + 𝑎1 𝑡 + 𝑎2 𝑡2 + 𝑎3 𝑡3 일 때, 적분 함수 𝐴 = 0 𝑡 𝑑𝑡  basis ∶ {1, 𝑡, 𝑡2, 𝑡3} 𝐴𝑃1 = 𝑡 = 𝑃2, 𝐴𝑃2 = 1 2 𝑡2 = 1 2 𝑃3, 𝐴𝑃3 = 1 3 𝑡3 = 1 3 𝑃4, 𝐴𝑃4 = 1 4 𝑡4 = 1 4 𝑃5 𝐴𝑃1 = 𝐴 1 0 0 0  0 1 0 0 0 , 𝐴𝑃2 = 𝐴 0 1 0 0  0 0 1 2 0 0 , 𝐴𝑃3 = 𝐴 0 0 1 0  0 0 0 1 3 0 , 𝐴𝑃4 = 𝐴 0 0 0 1  0 0 0 0 1 4 Linear Transformation Matrix A 𝐴 = 0 0 0 0 1 0 0 0 0 1 2 0 0 0 0 1 3 0 0 0 0 1 4 𝐴 𝑑𝑖𝑓𝑓 𝐴𝑖𝑛𝑡 = 0 1 0 0 0 0 0 2 0 0 0 0 0 3 0 0 0 0 0 4 0 0 0 0 1 0 0 0 0 1 2 0 0 0 0 1 3 0 0 0 0 1 4 = 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1 * 적분했다가 미분하면 같은 값 * 𝑨𝒊𝒏𝒕 𝑨 𝒅𝒊𝒇𝒇 ≠ 𝑰  미분했다가 적분하면 상수항이 생겨 같은 값 아님 * Differentiation is a left inverse of Integration.