In feature Scaling, we just use fit() method at train data. And not using in valid or test data.
Why we dont use mean and sd in test or valid data when we scaling test or valid data?
Asked
Active
Viewed 21 times
1
-
fit method is the training stage which is done only on the train set. Do you mean why do we scale only the train set by the mean and std? – theletz Jan 26 '20 at 05:23