contestada

if the performance measures from the training data are considerably higher than the values from the validation and test data, what could be the issue?