... it's usefull to split that dataset in 2 parts: a part for training and another for testing. The usual split is 80–20. Also, split it only if ...
確定! 回上一頁