New geometry links Naive Bayes classifiers to Shapley values

This paper introduces a novel geometric framework for understanding weighted Naive Bayes classifiers, particularly in the context of data streams with forgetting. Researchers developed a discriminative reformulation based on log-odds, which directly relates to classification decisions. This approach establishes a formal link between the model's induced geometry and analytical Shapley values, offering a new perspective on local explanations and predictive behavior. AI

IMPACT Introduces a novel geometric interpretation for weighted Naive Bayes classifiers, potentially improving model interpretability and explanation techniques.

RANK_REASON The cluster contains a research paper detailing a new methodological approach to classifier explanation. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• arXiv
• Hugging Face
• k-nearest neighbors classifier
• Shapley Values
• Weighted Naïve Bayes Classifier with Forgetting for Drifting Data Streams

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →