A Symbolic Approach to Explaining Bayesian Network Classifiers

TL;DR

This paper introduces a symbolic method for explaining Bayesian network classifiers by compiling them into decision functions, offering minimal feature-based explanations for classifications.

Contribution

It presents algorithms for generating two types of explanations and demonstrates their application to Naive and Latent-Tree Bayesian classifiers using ODDs.

Findings

Effective explanations for classifier decisions

Compilation of classifiers into ordered decision diagrams

Case studies validating the approach

Abstract

We propose an approach for explaining Bayesian network classifiers, which is based on compiling such classifiers into decision functions that have a tractable and symbolic form. We introduce two types of explanations for why a classifier may have classified an instance positively or negatively and suggest algorithms for computing these explanations. The first type of explanation identifies a minimal set of the currently active features that is responsible for the current classification, while the second type of explanation identifies a minimal set of features whose current state (active or not) is sufficient for the classification. We consider in particular the compilation of Naive and Latent-Tree Bayesian network classifiers into Ordered Decision Diagrams (ODDs), providing a context for evaluating our proposal using case studies and experiments based on classifiers from the literature.