Weighted Naive Bayes Classifier
An extension of the Naive Bayes classifier that assigns distinct weights to different features to improve classification performance. This variant maintains model interpretability while partially relaxing the strict conditional feature independence assumption.
Timeline (last 90 days)
The arXiv paper "From Log-Odds to Shapley Values: An Explanatory Geometry for the Weighted Naive Bayes Classifier" constructs a complete explanatory space for the weighted naive Bayes classifier.
The paper establishes formal mathematical connections among supervised representations, local explanations, and predictive behavior.
The paper provides an analytical form of Shapley values for weighted naive Bayes (a closed-form analytical solution requiring no sampling).
All Facts (3)
The arXiv paper "From Log-Odds to Shapley Values: An Explanatory Geometry for the Weighted Naive Bayes Classifier" constructs a complete explanatory space for the weighted naive Bayes classifier.
50%UnverifiedThe paper establishes formal mathematical connections among supervised representations, local explanations, and predictive behavior.
50%UnverifiedThe paper provides an analytical form of Shapley values for weighted naive Bayes (a closed-form analytical solution requiring no sampling).
50%