Prev: reset the contents of an edit box
Next: How to build a standalone application ?
From: corne.marais.4062 on 6 May 2008 03:46
I am developing a perceptron neural network with the following code:

%crossvalidation
blocksize = floor(nb_data/K);
for k = 1:K
if k==K
train=[1:blocksize*(k-1)];
validation=[blocksize*(k-1)+1 : nb_data];
else
train=[1:blocksize*(k-1) blocksize*k+1:nb_data];
validation=[blocksize*(k-1)+1 : blocksize*k];
end

%preprocess
[pxn meanx stdx]=prestd(px);

pxnt = pxn(:,train);
pxnv = pxn(:,validation);
pynewt = pynew(:,train);
pynewv = pynew(:,validation);

%backpropogation neural network
net = newff(pxn,pynew, 10,{'tansig','tansig'})

%train
net = train(net,pxnt,pynewt);

%test data
ypred = sim(net,pxnv);

end

The error message appears just after the network's data is given on my
workspace, I can't seem to find the problem even when implementing
dbstop error. I'm using matlab 7.5.0 (R2007b)
 | 
Pages: 1
Prev: reset the contents of an edit box
Next: How to build a standalone application ?