it is unfortunately not clear how the validation split is defined for each dataset. As the list of fields only indicate 'test_split'. Are we to find this information on OpenML website? Also, there appears no information on how to recreate the train/val/test splits.
it is unfortunately not clear how the validation split is defined for each dataset. As the list of fields only indicate 'test_split'. Are we to find this information on OpenML website? Also, there appears no information on how to recreate the train/val/test splits.