Tell me more ×
Cross Validated is a question and answer site for statisticians, data analysts, data miners and data visualization experts. It's 100% free, no registration required.
up vote 0 down vote favorite

I was wondering what are the implications of using a multi-class Naive Bayes versus a 2 class Naive Bayes (for one against everything).

Which technique performs better?

I've previously came across an instance where 2 class LDA classifier performed better than a multi-class LDA. Hence, this question.

I'm currently using Scikit-Learn machine learning package in python to use Naive Bayes classifier.

share|improve this question
1  
What is your problem? I don't see how you can compare two-class solutions to three classes solutions if the objective is to classify as correctly as possible an object with a given feature vector into the correct class when three are specified. Clearly a classifier using only information from classes 1 and 2 which may be able to correctly classify objects into class 1 and class 2 has no capability of clasifying correctly objects into class 3 as it has no decision boundary for class three. The rule is not equipped to classify into a category it has not created a region for. – Michael Chernick May 15 '12 at 21:47
I agree with Michael, how many classes you have ? – steffen May 16 '12 at 13:43
@steffen: I'm working with a 8 class classification problem. – raul_w May 16 '12 at 14:30
I think this could be useful. – kanzen_master Oct 12 '12 at 9:58

Know someone who can answer? Share a link to this question via email, Google+, Twitter, or Facebook.

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.