Limitations of Front-To-End Bidirectional Heuristic Search

Authors

  • Joseph Barker University of California, Los Angeles
  • Richard Korf University of California, Los Angeles

DOI:

https://doi.org/10.1609/aaai.v29i1.9374

Keywords:

Bidirectional Search, Pathfinding, Heuristic Search

Abstract

We present an intuitive explanation for the limited effectiveness of front-to-end bidirectional heuristic search, supported with extensive evidence from many commonly-studied domains. While previous work has proved the limitations of specific algorithms, we show that any front-to-end bidirectional heuristic search algorithm will likely be dominated by unidirectional heuristic search or bidirectional brute-force search. We also demonstrate a pathological case where bidirectional heuristic search is the dominant algorithm, so a stronger claim cannot be made. Finally, we show that on the four-peg Towers Of Hanoi with arbitrary start and goal states, bidirectional brute-force search outperforms unidirectional heuristic search using pattern-database heuristics.

Downloads

Published

2015-02-16

How to Cite

Barker, J., & Korf, R. (2015). Limitations of Front-To-End Bidirectional Heuristic Search. Proceedings of the AAAI Conference on Artificial Intelligence, 29(1). https://doi.org/10.1609/aaai.v29i1.9374

Issue

Section

AAAI Technical Track: Heuristic Search and Optimization