Naive Bayes assumes that predictors are independent. Though this assumption is quite powerful, in some scenarios it fails miserably . So are there any implementations of non Naive Bayes in Python ? Are there any issues that prohibit implementing non Naive Bayes Classifier.
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
A non-naive Bayes classifier is doable if one knows which specific features depend on each other, or if there are very few features. In any other case, the exponential number of dependencies requires an extremely large volume of instances in order for the classifier not to overfit.
The independence assumption in Naive Bayes is what makes the classifier able to generalize. The goal of a classifier is not to collect all the possible details in the data, it's to find the right balance between simplifying the data too much (underfitting) and not enough (overfitting).
