What does normalize=True in linear regression do? I searched online some of them answered it does not affect at all or only predictors are normalized. If the outcome is not affected why do we use it?
The prediction outcome shiuld not be afected by normalozation i.e. you model should right outcome in any case hopefully.
What normalization of features does is decrease and center their values so the model can rapidly converge to a good minima without much oscillations.