How to validate NARX network

2 visualizaciones (últimos 30 días)
Sara
Sara el 2 de Abr. de 2020
Hello,
I trained a NARX model with 5 neurons and 2 time delays. after training, I validated my model on unseen data using the following code. However, the error is very close to zero (MAE:0.07 RMSE: 0.19). I also attached the figure, showing the real data and estimated data. I think something is wrong with the code that I use for validation. Could you please help me? I think I can't use this result in my paper since it seem like a fake result.
Xtest1 = tonndata(testinput1,false,false);
Ttest1 = tonndata(testtarget1,false,false);
[xtest1,xitest1,aitest1,ttest1] = preparets(net,Xtest1,{},Ttest1);
outputtest1 = finalnet(xtest1,xitest1,aitest1);
errortest1 = gsubtract(ttest1,outputtest1);
rmse1= sqrt(mean((cell2mat(errortest1))'.^2))
MAE1= abs(mean(cell2mat(errortest1))')

Respuestas (0)

Categorías

Más información sobre Sequence and Numeric Feature Data Workflows en Help Center y File Exchange.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by