Learning Bayesian Nets that Perform Well

TL;DR

This paper addresses the challenge of learning Bayesian networks optimized for query-answering accuracy, highlighting the increased complexity compared to traditional likelihood-based methods.

Contribution

It introduces a focus on performance criteria related to query accuracy and demonstrates the computational difficulty of optimizing Bayesian nets for this purpose.

Findings

Optimizing BNs for query accuracy is more complex than likelihood-based methods.

Many aspects of this task are computationally harder than standard learning.

The paper highlights the gap between traditional learning objectives and actual query performance.

Abstract

A Bayesian net (BN) is more than a succinct way to encode a probabilistic distribution; it also corresponds to a function used to answer queries. A BN can therefore be evaluated by the accuracy of the answers it returns. Many algorithms for learning BNs, however, attempt to optimize another criterion (usually likelihood, possibly augmented with a regularizing term), which is independent of the distribution of queries that are posed. This paper takes the "performance criteria" seriously, and considers the challenge of computing the BN whose performance - read "accuracy over the distribution of queries" - is optimal. We show that many aspects of this learning task are more difficult than the corresponding subtasks in the standard model.