Description

Naive bayes in sklearn seems do not support additive smoothing only.
Can we make some change to make it support good-turing smoothing or other smoothing and how?
In some experiment, good-turing smoothing is better than additive smoothing.