矩阵范数和矩阵求导


今天发现两篇宝藏文章,关于矩阵范数和矩阵求导的,转载收藏一下。感谢大佬们的分享!{抱拳}
矩阵范数 转载自:https://blog.csdn.net/sylar49/article/details/77510160

今天看了半天强化学习,看得很不开心。。。因为一直处于懵圈状态。。。
于是乎不想看了,稍微总结一下矩阵范数的求解来放松一下身心吧~

这里总结的矩阵范数主要是F范数、1范数、2范数、核范数以及全变分TV范数与1、2的搭配

http://blog.csdn.net/txwh0820/article/details/46392293

矩阵求导 转载自:https://blog.csdn.net/txwh0820/article/details/46392293

矩阵的迹求导法则
矩阵的迹求导法则
矩阵的frobenius范数及其求偏导法则-1
矩阵的frobenius范数及其求偏导法则-2

1. 复杂矩阵问题求导方法:可以从小到大,从scalar到vector再到matrix
3
4
2. x is a column vector, A is a matrix

d(Ax)/dx=A" role="presentation" style="position: relative">d(A?x)/dx=A
d(xTA)/dxT=A" role="presentation" style="position: relative">d(xT?A)/dxT=A
d(xTA)/dx=AT" role="presentation" style="position: relative">d(xT?A)/dx=AT
d(xTAx)/dx=xT(AT+A)" role="presentation" style="position: relative">d(xT?A?x)/dx=xT(AT+A)

3. Practice:
5
4. 矩阵求导计算法则
求导公式(撇号为转置):
Y = A * X –> DY/DX = A’
Y = X * A –> DY/DX = A
Y = A’ * X * B –> DY/DX = A * B’
Y = A’ * X’ * B –> DY/DX = B * A’
乘积的导数:
d(f*g)/dx=(df’/dx)g+(dg/dx)f’

http://www.psi.toronto.edu/matrix/intro.html#Intro
http://www.psi.toronto.edu/matrix/calculus.html
http://www.stanford.edu/~dattorro/matrixcalc.pdf
http://www.colorado.edu/engineering/CAS/courses.d/IFEM.d/IFEM.AppD.d/IFEM.AppD.pdf
http://www4.ncsu.edu/~pfackler/MatCalc.pdf
http://center.uvt.nl/staff/magnus/wip12.pdf