행과 열의 크기가 같은 행렬을 Square Matrix 라고 합니다.
예제: \[ \begin{pmatrix} a_{11} & a_{12} & a_{13} \\ a_{21} & a_{22} & a_{23} \\ a_{31} & a_{32} & a_{33} \end{pmatrix} \]
어떤 행렬 \(A\)에 대한 Transpose Matrix 을 \(A^T\) 라고 합니다. \(A^T\)는 \(A\)의 각 행에 대응 되는 원소는 \(A^T\)의 열이 되며 \(A\)의 각 열에 대응 되는 원소는 \(A^T\)의 행에 대응 되는 것으로 정의 됩니다.
예제: 1. 번 예제의 Square Matrix의 Transpose Matrix는 다음과 같습니다. \[ \begin{pmatrix} a_{11} & a_{21} & a_{31} \\ a_{12} & a_{22} & a_{32} \\ a_{13} & a_{23} & a_{33} \end{pmatrix}\]
n by 1 크기의 행렬 \(Y\)에(벡터 Y) 대한 Transpose Matrix를 구하면 1 by n 크기의 행 벡터가 됩니다.
다음과 같이 주어진 회귀모델 \[ Y_i = E[Y_i] + \epsilon_i, i = 1,...,n \] 에 대한 행렬 표기법 \[ \begin{pmatrix} Y_1 \\ Y_2 \\ \vdots \\ Y_n \\ \end{pmatrix} = \begin{pmatrix} E[Y_1] \\ E[Y_2] \\ \vdots \\ E[Y_n] \\ \end{pmatrix} + \begin{pmatrix} \epsilon_1 \\ \epsilon_2 \\ \vdots \\ \epsilon_n \\ \end{pmatrix} = \begin{pmatrix} E[Y_1] + \epsilon_1 \\ E[Y_2] + \epsilon_2 \\ \vdots \\ E[Y_n] + \epsilon_n \\ \end{pmatrix} \]
\[ \underset{r\times s}{AB} = \left [ \sum_{k=1}^c a_{ik}b_{kj} \right ] \,\,\,\, i = 1,...,r; \,\, j = 1,...,s \]
\[ Y^TY = \begin{pmatrix} Y_1 & Y_2 & \cdots & Y_n \end{pmatrix} \begin{pmatrix} Y_1 \\ Y_2 \\ \vdots \\ Y_n \\ \end{pmatrix} = \begin{pmatrix} Y_1^2 & Y_2^2 & \cdots & Y_n^2 \end{pmatrix} = \left [ \sum_{k=1}^n Y_i^2 \right ] \]
\[ \underset{n \times 1}{1} = \begin{pmatrix} 1 \\ 1 \\ \vdots \\ 1 \\ \end{pmatrix} \]
\[ \underset{n \times n}{J} = \begin{pmatrix} 1 & \cdots & 1 \\ \vdots & & \vdots \\ 1 & \cdots & 1 \end{pmatrix} \]
\[ \underset{1 \times 1}{1^T1} = \begin{pmatrix} 1 & 1 & \cdots & 1 \end{pmatrix} \begin{pmatrix} 1 \\ 1 \\ \vdots \\ 1 \\ \end{pmatrix} = [n] = n = \left [ \sum_{k=1}^n 1^2 \right ]\]
\[ \begin{pmatrix} 1 \\ 1 \\ \vdots \\ 1 \\ \end{pmatrix} \begin{pmatrix} 1 & 1 & \cdots & 1 \end{pmatrix} = \begin{pmatrix} 1 & \cdots & 1 \\ \vdots & & \vdots \\ 1 & \cdots & 1 \end{pmatrix} = \underset{n \times n}{J} = \left [ \sum_{k=1}^1 a_{ik}b_{kj} \right ] \,\, i = 1,...,n; \,\, j = 1,...,n \] 정상적인 공유기능을 활용한 공유가 아닌 화면 캡쳐 및 강제로 복사 및 붙여 넣기 식의 공유를 일체 하지 말아 주세요. 출처를 밝힌다고 해도 화면 캡쳐 및 강제로 복사 및 붙여 넣기 식의 공유를 절대 금지합니다.