I have data sets with train, test and validate and want to use the MLPClassifier in sklearn. However are there any examples that can be used for this since Im not sure how to use all of these separate datasets? If running tuning by eg GridSearchCV only one training dataset is used which is split by the GridSearchCV itself (training and test). How can I train, test and validate my MLPClassifier using my different datasets? How is it usually done? Also it would be nice to have plot for accuacy, loss versus epoch during train and test during.
I have tried gridsearch with training dataset but how should I use the test and validation dataset? If I use my best gridserach parameters and check with test parameters it seems that I have no use of my validation data set or how is this commonly used, exemplify with code.