Normal Bayes implementation in OpenCV

Question

I'm a newbie to Machine Learning. I have a question about how Normal Bayes is implemented in OpenCV.

I have a mis-understanding regarding the terms Normal Bayes and Naive Bayes.

This site tells that Normal Bayes and Naive Bayes mean the same.

The NormalBayes documentation on OpenCV website specifies that the features are Normally distributed and not necessarily independent.

The wikipedia article on Naive Bayes classifier tells us that it is assumed that features are independent. Therefore, Covariance Matrix need not be determined.

However, when I look at the source of the implementation of Normal Bayes classifier, it does calculate Covariance Matrix.

I also found a similar question over here which wasn't answered.

Am I missing something here? or is it that Normal Bayes classifier in OpenCV is not a standard Naive Bayes classifier?

Moacir Ponti · Accepted Answer · 2012-04-05 14:12:13Z

3

Theoretically, the Naive Bayes model assumes "complete independence between causes of an eﬀect", while the Normal model assumes that "feature vectors from each class are normally distributed (though, not necessarily independently distributed)". Note that both uses mean vectors and covariance matrices, however, the model assumption is different.

In OpenCV "data distribution function is assumed to be a Gaussian mixture, one component per class" and the model does not made an assumption regarding independence of such classes.

answered Apr 5, 2012 at 14:12

Moacir Ponti

5915 silver badges14 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

Normal Bayes implementation in OpenCV

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related