Mini-Bucket Heuristics for Improved Search

TL;DR

This paper extends the use of mini-bucket heuristics for search algorithms, demonstrating improved performance in best-first search over branch-and-bound in coding and medical diagnosis problems.

Contribution

It introduces an extension of mini-bucket heuristics evaluation within best-first search, showing how to balance preprocessing and search efficiency.

Findings

Best-first search outperforms branch-and-bound with good heuristics

Mini-bucket heuristics enable controlled tradeoff between preprocessing and search

Effective search scheme demonstrated on coding and medical diagnosis problems

Abstract

The paper is a second in a series of two papers evaluating the power of a new scheme that generates search heuristics mechanically. The heuristics are extracted from an approximation scheme called mini-bucket elimination that was recently introduced. The first paper introduced the idea and evaluated it within Branch-and-Bound search. In the current paper the idea is further extended and evaluated within Best-First search. The resulting algorithms are compared on coding and medical diagnosis problems, using varying strength of the mini-bucket heuristics. Our results demonstrate an effective search scheme that permits controlled tradeoff between preprocessing (for heuristic generation) and search. Best-first search is shown to outperform Branch-and-Bound, when supplied with good heuristics, and sufficient memory space.