Github user sethah commented on the issue:
https://github.com/apache/spark/pull/15324
Was there some discussion as to whether GaussianNB should be part of the
NaiveBayes estimator or its own estimator? It seems the semantics are different
enough between multinomial NB and Gaussian NB to at least warrant discussion.
The meaning `theta` matrix in Gaussian NB vs multinomial is very different in
this patch - one is matrix of Gaussian distribution paramters, and the other is
a matrix of class conditional probabilities. Also, some params only apply to
one and not the other. My apologies if I have missed this conversation
somewhere. Thanks!
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
