net=newp([-2 5;-5 2],1)
net.inputweights{1,1}.initFcn="rand";
net.biases{1}.initFcn="rand";
net=init(net);
wts=net.iw{1,1}
wts =
Empty matrix: 1-by-0
>> bias=net.b{1}
bias =
0.9501
为什么wts =Empty matrix: 1-by-0 而不是随机的数那
Create By Any-Extract(WL-AE) 数据挖掘研究院
