This question is closed. Reopen it to edit or answer.

Widrow-Hoff delta rule method with linear layer (adaline)

1 view (last 30 days)
HASAN AL-KAF on 27 Sep 2017
Closed: Stephen23 on 27 Sep 2017
i want to use learnwb which Widrow-Hoff delta rule (also known as least mean square (LMS) algorithm) and updatesits weights and bias when a training sample is presented. this is my code x=401*113 y=1*113 net = linearlayer(0,0.001); net = configure(net,x,y); net.trainFcn = 'trainb'; net.trainParam.epochs = 788; net=train(net,x,y); outputs=net(x); e=output-y ee=e.^2 eee=sum(ee) rmse=sqrt(mean(eee)); zz= postreg (outputs , y) but my error is root mean square error is very high ? could you please give my some suggestion ? is it my training method correct ?
  1 Comment
Stephen23 on 27 Sep 2017

Answers (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!