Multivariate Bernoulli vs Multinomial Naive Bayes

bhaiya , when to use multivariate bernouli naiyes bayes and when to use multinomail naive bayes?

Hey aman,
Three different types of Naive bayes are:

  • Multivariate Bernaulli
  • Mutinomail NB
  • Gaussian NB

When we have continuous values in features. Then it is preferred to use Gaussian NB.
When we have discrete features then either Multivariate or Multinomial.
Further when we have feature values as discrete but Boolean only, either True/Flase 1 / 0 only 2 types of values possibles in feature, in that case Multivariate Bernaulli is preferred,
And if features are discrete but can have multiple values then, go for Multinomial NB.

I hope use cases of all three are clear to you now

Thanks :slight_smile:

Thanks bhaiya:grinning: