Dear Bryan Escoto

To check the accuracy of the artificial neural network model in MATLAB, you can check the Regression value, MSE, and Error histogram.

high R, less MSE, Fewer errors will be good for your network.

Also, check this video:

https://kr.mathworks.com/videos/getting-started-with-neural-networks-using-matlab-1591081815576.html

Right now, I am working in python. if you are interested in python then I can give an idea about the accuracy calculation.

Dear Mr. Escato,

You may use "help nnperformance" command for listing possible functions and their definitions.

Mean Squared Error [MSE] is the most common error performance function and can be chosen with;

net.performFcn = 'mse';

Additionally, you may check performance of your network by plotting errors. For this purpose "help nnplot" will list all possible options.

net.plotFcns = {'plotperform', 'plottrainstate', 'ploterrhist', 'plotregression', 'plotfit'};

Regression Plots also shows R values that provides an easier way to measure accuracy of your network.

Dear Bryan Escoto

To evaluate the performance of your model during validating or testing step, you have to calculate determination of coefficient and mean squared error. Then, there are possibility of encountering networks with lower R^2 and lower MSE or higher R^2 and higher MSE. To provide an approach to make distinguish between them, we used a goodness function in our recently published work. Please check below article for more details.

Article Application of artificial neural network and full factorial ...

Good luck.

[YPred,scores] = classify(netTransfer,augimdsValidation);

Please let me how to calculate the accuracy of artificial neural network modeled in R?