How to interpret this correlation graph in Neural Networks?

I'm modeling a NARX with 1 target series and 5 additional variables (size of sample 162 and delay=12). I've used 90 hidden neurons and bayesian regularization for the training.

What does it mean this graph?

Thanks in advance for your help.

Answer 1

[Greg Heath]

The magnitude of the correlations between the inputs and errors are large compared to the red line of significance. Therefore you should be able to significantly improve your design.

Hope this helps.

Thank you for formally accepting my answer

Greg

[Random_Noise]

Dear Greg,

Thank for your answer. Could you give some guidance in order to improve the design, please?.

Best, José

[Greg Heath]

I have many posts on NARXNET in BOTH the NEWSGROUP and ANSWERS

Just search using

 greg narxnet

Check the most recent posts first.

Hope this helps,

Greg

[Random_Noise]

I'll check that Greg. Many thanks!

Best, J.