I don’t find train-test distinctions particularly essential here because our method is unsupervised
If I recall correctly, most unsupervised learning papers do have a test set. Perhaps the fact that the train and test are different kind of shows why you need a test set in the first place.
If I recall correctly, most unsupervised learning papers do have a test set. Perhaps the fact that the train and test are different kind of shows why you need a test set in the first place.