Systematic vs. Non-systematic Algorithms for Solving the MPE Task

TL;DR

This paper compares systematic Branch and Bound algorithms, BBBT and BBMB, with local search methods for the MPE task in Bayesian Networks, showing the systematic algorithms outperform local search especially in larger domains.

Contribution

It demonstrates that BBBT and BBMB are the most effective algorithms for MPE, outperforming local search methods, especially as domain sizes grow.

Findings

BBBT/BBMB outperform local search algorithms for MPE.

Systematic algorithms are superior to local search in larger domains.

BBBT/BBMB are the best known algorithms for MPE.

Abstract

The paper continues the study of partitioning based inference of heuristics for search in the context of solving the Most Probable Explanation task in Bayesian Networks. We compare two systematic Branch and Bound search algorithms, BBBT (for which the heuristic information is constructed during search and allows dynamic variable/value ordering) and its predecessor BBMB (for which the heuristic information is pre-compiled), against a number of popular local search algorithms for the MPE problem. We show empirically that, when viewed as approximation schemes, BBBT/BBMB are superior to all of these best known SLS algorithms, especially when the domain sizes increase beyond 2. This is in contrast with the performance of SLS vs. systematic search on CSP/SAT problems, where SLS often significantly outperforms systematic algorithms. As far as we know, BBBT/BBMB are currently the best performing algorithms for solving the MPE task.