Hello everyone) help please deal with NA in Matlab. There is an input from 66 neurons, and an output with one neuron. Such examples 600. everything is learning normally.

net=newff([-1 1],[5 1],{'tansig' 'purelin'}); net.trainParam.epochs=1000; net.inputs{1}.size=66; net=train(net,input,output);

I want to know the answer to the test data, I write to the net command line ({vector of 66 numbers}), displays 66 values, and not one, as it should be on output. What to do? How to find out the answer?

  • If you are given an exhaustive answer, mark it as correct (a daw opposite the selected answer). - Nicolas Chabanovsky ♦

1 answer 1

Network modeling (using the network to solve the problem).

In order to apply a trained network to data processing, you need to use the sim function:

 Y=sim(net,p);

where p is the set of input vectors, Y is the result of the analysis in the form of a set of output vectors.