Weak Proof-Number Search

Authors:
Toru Ueda;Tsuyoshi Hashimoto;Junichi Hashimoto;Hiroyuki Iida
Affiliations:
School of Information Science, Japan Advanced Institute of Science and Technology, Ishikawa, Japan 923-1292;School of Information Science, Japan Advanced Institute of Science and Technology, Ishikawa, Japan 923-1292;School of Information Science, Japan Advanced Institute of Science and Technology, Ishikawa, Japan 923-1292;School of Information Science, Japan Advanced Institute of Science and Technology, Ishikawa, Japan 923-1292
Venue:
CG '08 Proceedings of the 6th international conference on Computers and Games
Year:
2008

Citing 7
Cited 0

Conspiracy numbers for min-max search

Artificial Intelligence
Conspiracy numbers

Artificial Intelligence - Special issue on computer chess
Linear-space best-first search

Artificial Intelligence
Proof-number search

Artificial Intelligence
The graph-history interaction: on ignoring position history

ACM '85 Proceedings of the 1985 ACM annual conference on The range of computing : mid-80's perspective: mid-80's perspective
The PN -search algorithm: application to tsume-shogi

Artificial Intelligence - Special issue on heuristic search in artificial intelligence
A Solution to the GHI Problem for Best-First Search

CG '98 Proceedings of the First International Conference on Computers and Games

Quantified Score

Hi-index	0.00

Visualization

Abstract

The paper concerns an AND/OR-tree search algorithm to solve hard problems. Proof-number search is a well-known powerful search algorithm for that purpose. Its depth-first variants such as PN*, PDS, and df-pn work very well, in particular in the domain of shogi mating problems. However, there are still possible drawbacks. The most prevailing one is the double-counting problem. To handle this problem the paper proposes a new search idea using proof number and branching factor as search estimators. We call the new method Weak Proof-Number Search. The experiments performed in the domain of shogi and Othello show that the proposed search algorithm is potentially more powerful than the original proof-number search or its depth-first variants.