On Robust Trimming of Bayesian Network Classifiers

TL;DR

This paper introduces a method for trimming Bayesian network classifiers by removing costly features while maintaining robustness and accuracy, using a new closeness metric and an optimal feature selection algorithm.

Contribution

It proposes a novel ECA metric for classifier similarity and an optimal trimming algorithm that balances feature reduction with classifier robustness.

Findings

The trimming algorithm effectively reduces features while preserving classification behavior.

Experimental results show improved robustness and maintained accuracy after trimming.

The approach offers a computationally feasible way to optimize feature subsets in Bayesian classifiers.

Abstract

This paper considers the problem of removing costly features from a Bayesian network classifier. We want the classifier to be robust to these changes, and maintain its classification behavior. To this end, we propose a closeness metric between Bayesian classifiers, called the expected classification agreement (ECA). Our corresponding trimming algorithm finds an optimal subset of features and a new classification threshold that maximize the expected agreement, subject to a budgetary constraint. It utilizes new theoretical insights to perform branch-and-bound search in the space of feature sets, while computing bounds on the ECA. Our experiments investigate both the runtime cost of trimming and its effect on the robustness and accuracy of the final classifier.