Naive Bayes For Regression

Question

I was wondering, if I can apply naive bayes, to a regression problem and how will it be done. I have 4096 image features and 384 text features and, it won't be very bad if I assume independence between them. Can anyone tell me how to proceed?

Answer 1

Naive bayes is used for strings and numbers(categorically) it can be used for classification so it can be either 1 or 0 nothing in between like 0.5 (regression)

Even if we force naive bayes and tweak it a little bit for regression the result is disappointing; A team experimented with this and achieve not so good results.

Also in wikipedia naivebayes has closeness to logistic regression.

Relation to logistic regression: naive Bayes classifier can be considered a way of fitting a probability model that optimizes the joint likelihood p(C , x), while logistic regression fits the same probability model to optimize the conditional p(C | x).

So now you have two choices, tweak naive bayes formula or use logistic regression.

I say lets use logistic regression instead of reinventing the wheel.

References:

Wikipedia: https://en.wikipedia.org/wiki/Naive_Bayes_classifier#Relation_to_logistic_regression

Naive Bayes Regression Experiment: https://link.springer.com/content/pdf/10.1023%2FA%3A1007670802811.pdf

Answer 2

Naive bayes doesn't make sense to me as a regression algorithm. Random forest regression might be a better fit for your problem. It should be able to handle mixed text and image features.