Neural networks toolbox, error weights, get an error
1 view (last 30 days)
Show older comments
Hello!
I'm trying to run the following script
x=eye(12);
t = [0 1 1 0 1 0 0 0 1 1 0 1];
ew = (t==0)*0.5 + (t==1);
net = feedforwardnet(10);
net = configure(net,x,t);
net = train(net,x,t,[],[],ew);
And i got an error message ??? Error using ==> trainlm at 109 Inputs and input states have different numbers of samples.
Error in ==> network.train at 107 [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);
Error in ==> Untitled at 7 net = train(net,x,t,[],[],ew);
Without the ew parameter it works pretty good:
net = train(net,x,t);
The question is, how can I properly set error weights for targets? I use MATLAB 2010b.
In fact I tried to run an example from here http://www.mathworks.de/matlabcentral/answers/2868-neural-network-performance-function-weighted-sse-and-false-alarms
Thank you in advance, Leonid
1 Comment
PALASH MONDAL
on 9 Mar 2018
My dear as your mention code above is now perfectly running on Matlab 2017a without any modification. There is no error as you mention as line wise.
Answers (2)
Francois
on 14 Jul 2011
Contacted Mathworks,
this is the right code
[net,tr] = train(net,x,t,{},{},EW);
I used it, and it works great!
2 Comments
Francois
on 7 Jul 2011
I have the same problem,
[net,tr] = train(net,x,t,[],[],EW);
??? Error using ==> trainlm at 109
Inputs and input states have different numbers of samples.
Error in ==> network.train at 107
[net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);
the problem is not with EW but with Xi and Ai, and this is not easy to write. If someone has done it before an example would be appreciated!
1 Comment
See Also
Categories
Find more on Sequence and Numeric Feature Data Workflows in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!