Can we report our results in the journal paper when we don't use the validation set? Or It is better to ask when we can use train and test sets without validation? Can you give me a reference? My dataset doesn't have enough data to split that train, valid, and test. So, I want to train my model without a validation set. But I need a reference. Can you help me? Thank you
2.1m questions
2.1m answers
60 comments
57.0k users