A Symbolic Approach to Explaining Bayesian Network Classifiers
This work addresses the interpretability issue for users of Bayesian network classifiers, though it appears incremental as it builds on existing compilation techniques.
The authors tackled the problem of explaining Bayesian network classifiers by compiling them into symbolic decision functions, specifically Ordered Decision Diagrams, to generate minimal feature-based explanations for classifications.
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.