I have independent training, validation and test sets, all from different years. So far i have trained my models and compared their results on the validation set. Im wondering if it makes sense to resample my training set, training my models in different subsamples and comparing the results with the average of the metrics in the validation set. My sets have around 30k instances. Am i being overzealous?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.