MATLAB: Feedback Delay of a Layer Recurrent Neural Network with multiple Outputs

layer recurrentlayrecnetmachine learningneural networks

My Neural Network has multiple outputs and I'm trying to find the right Feedback delay so I'm computing the correlation of my output vector, but I don't know if I have to calculate the autocorrelation : nncorr(Output1,Output1) or do I have to calculate the crosscorrelation: nncorr(Output1,Output2) etc … because it's a neuralnet with multiple outputs?

Best Answer

The only feedback to the input is from the hidden layer.

However, given the input, there is no way to estimate that signal. Ttherefore, there is no apriori way to esrimate the significant crosscorrelation delays.

Related Solutions

MATLAB: How to use neural network for multi step ahead prediction

Increase the number of feedback delays.

Hope this helps

Thank you for formally accepting my answer

Greg

MATLAB: 20 Days ahead forecasts using NAR

Try

 FD = 1:20;
 net = narnet(FD,H=10);
 net.divideFcn = 'divideblock';  % Never use 'dividerand' for timeseries.

For details see

 help narnet
 doc narnet

and search the NEWSGROUP and ANSWERS using

 greg narnet