CODEWITHSUNDEEP
classificationmultilabel-classification
mahsa
mahsa

Reputation: 17

multi class naive bayes

I don't have much experience in the data mining. My question is about the multi-class naive bayes classification . I need a simple example in this field if I want to implement this algorithm on 3 lable input and 3 output classes. My other question is if i use 2-class naive bayes (for example :positive and negative) and have two output class ,but i nead 4 output class(very positive,positive,negative,very negative) , can i use a threshold for a positive result that if output is greater than the threshold value it means it is in very positive class?

Upvotes: 0

Views: 8467

Answers (1)

badc0re
badc0re

Reputation: 3523

You can have threshold for each class label, if you use Naive Bayes that will depend on the probability of the class when you classify. Also it doesn't matter how many classes or features you are using as long as you distinct the features used by each class. You can read this article to clear things up.

Upvotes: 2

Related Questions