(三)Multi-class Classification and Neural Networks[多分类问题和神经网络]
这次打算以代码为主线,适当补充。问题:手写数字识别。方法一:逻辑回归forc=1:num_labelsinitial_theta=zeros(n+1,1);%Setoptionsforfminuncoptions=optimset('GradObj','on','MaxIter',50);[theta]=fmincg(@(t)(lrCostFunction(t,X,(y==c),lambda)),