It's common in machine learning to train on 70% of non-validation data, testing on the remaining 30%. I'm not sure whether the motive for this is theoretical or "empirically this works well", although my reading suggests the latter. Where was this ratio first justified either way?



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Browse other questions tagged or ask your own question.