CODEWITHSUNDEEP
machine-learningdata-mining
v09
v09

Reputation: 890

naive bayes classifier on continuous data (real valued data)

I am new to machine learing/Data mining field.

I was asked question sometimes back,

Does naive bayes work with continuous features (real valued data)? If I have a feature set (f1,f2,f3........) all real valued data points. Does naive bayes work ?

The answer for the question was. It does not.

Can somebody explain why it does not work on real valued data?

What should be done to classify (real valued data)?

Upvotes: 0

Views: 631

Answers (1)

lejlot
lejlot

Reputation: 66835

The answer was wrong. Naive Bayes can work on arbitrary data, simply you have to change the method of probability estimation. One of the simplest options is so called "gaussian naive bayes" which assumes gaussian distribution of features, which makes working with arbitrary real valued features possible.

Upvotes: 1

Related Questions