Is it a fact that the training, validation, and test splits of the entire data set were done completely randomly? In other words, can we assume that the training, validation, and test data sets come from the same distribution?
1 Like
Hi @mlguru,
Yes, you can assume that that the training, validation, and test sets come from the same distribution.
1 Like