How can calculate ( MSE , NMSE , Correlation Coefficient ) for the output result of neural network ?

I want to calculate ( Mean Square Error , Normalized Mean Square Error , Correlation Coefficient ) for the output result of neural network using matlab code ...

the Data are a group of different images, its features were extracted using PCA and then entered into the Nueral Network.

How can I know about system performance ?

PLZ any help .. thanks

[yuval]

You should give some more information on what you're trying to do, including relevant bits of your code, because I don't see a way to help otherwise.

[Greg Heath]

In particular, you have not defined an output!

Greg

[Aya Ahmed]

yuval,, I want to make image classification in to three types of images

[Aya Ahmed]

Greg Heath the output is like that : -

three classes of images ,, i want to get there performance like Mean Square Error , Normalized Mean Square Error , Correlation Coefficient

Answer 1

[masoud sistaninejad]

inputs = x_test ; % your test data inputs

targets = y_test ; % your test data targets

P = net(inputs);

Error = targets - P;

MSE = mse(targets,P);

NMSE = (mean(error.^2)) / (mean(var(targets',1)));