Linear Algebra-Lecture04 矩阵A的LU分解
矩阵乘积的逆
若 A \bf A A都逆矩阵存在,则 A A − 1 = I = A − 1 A \bf A \bf A^{-1}=\bf I =\bf A^{-1} \bf A AA−1=I=A−1A
若 A \bf A A与 B \bf B B的逆矩阵都存在,则 ( A B ) − 1 = B − 1 A − 1 \bf {(AB)^{-1}}=\bf B^{-1}\bf A^{-1} (AB)−1=B−1A−1
解释: ( A B ) B − 1 A − 1 = A ( B B − 1 ) A − 1 = I \bf{(AB)}\bf B^{-1}A^{-1}=\bf{A(BB^{-1})A^{-1}}=\bf I (AB)B−1A−1=A(BB−1)A−1=I,因此 A B \bf {AB} AB的逆矩阵为 B − 1 A − 1 \bf B^{-1}\bf A^{-1} B−1A−1。
转置的逆
A A − 1 = I ⇒ ( A A − 1 ) T = I T ⇒ ( A − 1 ) T A T = I \bf A \bf A^{-1}=\bf I\qquad\Rightarrow\bf {(A A^{-1})^T}=\bold{I}^T\qquad\Rightarrow\bf {(A^{-1})^TA^T}=\bold{I} AA−1=I⇒(AA−1)T=IT⇒(A−1)TAT=I,因此, A T \bf A^T AT的逆为 ( A − 1 ) T \bf (A^{-1})^T (A−1)T,即 ( A T ) − 1 = ( A − 1 ) T \bf (A^T)^{-1}=\bf (A^{-1})^T (AT)−1=(A−1)T,换句话说,对于单个矩阵而言,转置和逆两种运算,其顺序可以颠倒。
LU分解
设 A = [ 2 1 8 7 ] \bf A=\begin{bmatrix}2&1\\8&7\end{bmatrix} A=[2817],用初等矩阵 E 21 = [ 1 0 − 4 1 ] \bf E_{21}=\begin{bmatrix}1&0\\-4&1\end{bmatrix} E21=[1−401]进行行变换,可得到上三角矩阵 U = [ 2 1 0 3 ] \bf U=\begin{bmatrix}2&1\\0&3\end{bmatrix} U=[2013].
E 21 A = U [ 1 0 − 4 1 ] [ 2 1 8 7 ] = [ 2 1 0 3 ] \begin{array}{cccc} \bf E_{21}&\bf A&=&\bf U\\ \begin{bmatrix}1&0\\-4&1\end{bmatrix}&\begin{bmatrix}2&1\\8&7\end{bmatrix}&=&\begin{bmatrix}2&1\\0&3\end{bmatrix} \end{array} E21