Topic: Artificial nueral network help
Detailed Question: I have configured a network and trained it with 4 inputs and 2 outputs. the results of the training R and Testing R developed by MATLAB is in the range of 0.85-0.95 the outputs is simulated and RMSE is calculated , the results in the range of( 0.3 - 0.4) but the absolute average error for the second output is high (10 %) and correlation factor for the second output showed very low results (0.002) and somepoints in negative while for the first output is high and higher than 0.9 if you can have a look on the MATLAB code to improve the prediction of 2nd output and explain how the RMSE is low and network training and testing R above 0.9 while the correlation factor for the 2nd output is very low.