Re:A practical explanation of a Naive Bayes classifier
by rgarreta on Friday July 28, 2017 @01:31PM
(#14875971)
I would add that another practical aspect about Naive Bayes classifiers is that you can make use of the conditional probabilities for each feature that contributes to the predictions. That gives you some introspection on how the model is working and it's useful when "debugging" classifiers by finding features that should/shouldn't be used.
Re:A practical explanation of a Naive Bayes classifier
by abhgh on Friday July 28, 2017 @03:14PM
(#14876728)
A side note here: the classification probabilities NB produces are not very accurate and usually need some correction using a process called "calibration".
https://monkeylearn.com/blog/how-to-create-text-classifiers-...