How to calculate mean square error for 3 input and 2 output neural network architecture? How to plot regression curve for predicted outputs vs target using ANN-PSO optimizatio

I have 3 inputs and 2 outputs.

I am using optimizing weights of ANN using PSO and predicting the 2 outputs.

How to calculate Mean Squared error for the two outputs? How to plot the predicted values of 2 ouputs vs target values?

[Sarmed Wahab]

MSE is between predicted and actual target values. So just use the MSE formula or mse function of MATLAB.

For plotting the predicted values against the actual values of output, you can use the scatter plot.

scatter(predicted_values, actual_values);
plot([0,1],[0,1]);

Or you can use the saved results to generate the regression plots

plotregression(trTarg_ANNPSO,trOut_ANNPSO,'Train',tsTarg_ANNPSO,tsOut_ANNPSO,'Testing',targets, net_ANNPSO(inputs),'All');

Where,

tr = training data proportion

ts= testing data proportion

Targ = target value

Out = predicted values