Abstract: In modeling, it is generally to divide the dataset into training and validation sub-datasets. Although it appears simple, how to divide the dataset is still somewhat debatable. Of various ...