MATLAB: How to test a neural network with an unknown input data set

how to test a neural network with an unknown input data set?

Hi,

I have used fitting tool of the neural network. After training (validation and testing)the network with the related input and target data sets, now all I need is to apply new input to the network and predict the values of corresponding outputs. I do not have target data set for new testing and I do not need to know the performance as well, just need to to know the predicted values. I need a function like PREDICT used in DECISION TREE classifier……

Best Answer

ynew = net(xnew)

Hope this helps.

Greg

Related Solutions

MATLAB: How to test a trained network with new data

 trueclass        = vec2ind(target)  % Unit vector target columns --> class indices
 [net tr output ] = train(net,input,target);
 assignedclass    = vec2ind(output)
 err              = assignedclas~=trueclass
 Nerr             = sum(err)
 newoutput        = net(newinput)
 newassignedclass = vec2ind(newoutput)

Thank you for formally accepting my answer

Greg

MATLAB: How to do testing and training using feed forward neural network

Use patternnet, not feedforwardnet. The target matrix should be a row vector containg zeros and ones indicating the true class of the corresponding input.

help patternnet

doc patternnet

Hope this helps.

Greg

Best Answer

Related Solutions

MATLAB: How to test a trained network with new data

MATLAB: How to do testing and training using feed forward neural network

Related Question