2.矩阵的迹&转置&对称矩阵
2.1 矩阵的迹
定义:
- \(n \times n\)矩阵主对角线上元素的总和称为\(矩阵的迹\)
- 矩阵X的迹记为\(tr(X)\)
示例:
设存在以下\(n \times n\)的矩阵:
\[X_{n \times n}=
\begin{bmatrix}
x_{11} & x_{12} & x_{13} & ... & x_{1n}\\
x_{21} & x_{22} & x_{23} & ... & x_{2n}\\
x_{31} & x_{32} & x_{33} & ... & x_{3n}\\
&&......\\
x_{n1} & x_{n2} & x_{n3} & ... & x_{nn}\\
\end{bmatrix}\\
\]
则:
\[tr(X)=x_{11}+x_{22}+x_{33}+...+x_{nn}=\sum_{i=0}^n x_{ii}
\]
矩阵的迹运算律
\[\tag {1}tr(A\cdot B)=tr(B\cdot A)
\]
证明:
\[设A\cdot B=C,B\cdot A=D,则有:\\
\]
\[tr(A\cdot B)=tr(C)=\sum_{i=1}^nc_{ii}=\sum_{i=1}^n\sum_{j=1}^na_{ij}b_{ji}\\
\]
\[tr(B\cdot A)=tr(D)=\sum_{i=1}^n\sum_{j=1}^nb_{ij}a_{ji}=\sum_{j=1}^n\sum_{i=1}^na_{ji}b_{ij}=\sum_{i=1}^n\sum_{j=1}^na_{ij}b_{ji}\\
\]
\[\Rightarrow tr(A\cdot B)=tr(B\cdot A)
\]
2.2 矩阵的转置
定义:
- 把一个矩阵的行换成同序数的列,得到一个新的矩阵。
- 矩阵\(X\)的转置记为\(X^T\)
性质:
- 若X为\(m \times n\)的矩阵,则\(X^T\)为\(n \times m\)的矩阵
- 设\(x\)为矩阵\(X\)中的元素,\(x^T\)为矩阵\(X^T\)中的元素,则\(x^T_{ij}=x_{ji}\)
示例:
设存在以下\(m \times n\)的矩阵:
\[X_{m \times n}=
\begin{bmatrix}
x_{11} & x_{12} & x_{13} & ... & x_{1n}\\
x_{21} & x_{22} & x_{23} & ... & x_{2n}\\
x_{31} & x_{32} & x_{33} & ... & x_{3n}\\
&&......\\
x_{m1} & x_{m2} & x_{m3} & ... & x_{mn}\\
\end{bmatrix}\\
\]
则:
\[X^T_{n \times m}=
\begin{bmatrix}
x_{11} & x_{21} & x_{31} & ... & x_{m1}\\
x_{12} & x_{22} & x_{32} & ... & x_{m2}\\
x_{13} & x_{23} & x_{33} & ... & x_{m3}\\
&&......\\
x_{1n} & x_{2n} & x_{3n} & ... & x_{mn}\\
\end{bmatrix}\\
\]
矩阵的转置运算律
\[\tag {1}(A^T)^T=A
\]
\[\tag {2}(A+B)^T=A^T+B^T
\]
\[\tag {3}(\lambda\cdot A)^T=\lambda\cdot A^T
\]
\[\tag {4}(A\cdot B)^T=B^T\cdot A^T
\]
对运算律(4)进行证明:
\[设存在矩阵A_{m\times t},矩阵B_{t\times n},且A_{m\times t}\cdot B_{t\times n}=C_{n\times n},则:\\
c_{ij}=\sum_{k=1}^ta_{ik}\cdot b_{kj}\\
\]
\[由矩阵转置的相关性质可知:\\c^T_{ij}=c_{ji}=\sum_{k=1}^ta_{jk}\cdot b_{ki}\\
\]
\[设B^T_{n \times s}\cdot A^T_{s \times n} =D_{n\times n},则:\\
d_{ij}=\sum_{k=1}^tb^T_{ik}\cdot a^T_{kj}=\sum_{k=1}^tb_{ki}\cdot a_{jk}=\sum_{k=1}^ta_{jk}\cdot b_{ki}=c^T_{ij}
\]
\[\Rightarrow (A\cdot B)^T=B^T \cdot A^T
\]