How to properly train the network MPL?

Dear all,

I have this code:

P = [c{1,:}];
T = [c{2,:}];
net1 = newff(P,T,20);
net1.trainparam.epochs = 100;
net = train(net1,P,T);
P = s{1,7};
Y = sim(net1,P);

After running the code, nothing happens. Just Comand Window reports name of file. Why not start training network? Does anyone have any idea? How to make this network trains?

Answer 1

[Greg Heath]

You seem to repeatedly make this same mistake:

 net = train(net1,P,T);

 Y = sim(net1,P);

 Now: There are 2 nets: net1 and net

 Question: which is trained and which is untrained?

PS: MLP(Multi-Layer-Perceptron), not MPL

[Veronika]

That's what I just don´t know. Because I have this piece of code created from Help in Matlab. I thought that this is the exact syntax. So I don´t know, which one is trained and which aren´t. I need to train the neural network with the given parameters (P,T) and then save this train netwok.