Computing Elo Ratings of Move Patterns in the Game of Go

Rémi Coulom 1
1 SEQUEL - Sequential Learning
LIFL - Laboratoire d'Informatique Fondamentale de Lille, Inria Lille - Nord Europe, LAGIS - Laboratoire d'Automatique, Génie Informatique et Signal
Abstract : Move patterns are an essential method to incorporate domain knowledge into Go-playing programs. This paper presents a new Bayesian technique for supervised learning of such patterns from game records, based on a generalization of Elo ratings. Each sample move in the training data is considered as a victory of a team of pattern features. Elo ratings of individual pattern features are computed from these victories, and can be used in previously unseen positions to compute a probability distribution over legal moves. In this approach, several pattern features may be combined, without an exponential cost in the number of features. Despite a very small number of training games (652), this algorithm outperforms most previous pattern-learning algorithms, both in terms of mean log-evidence (−2.69), and prediction rate (34.9%). A 19x19 Monte-Carlo program improved with these patterns reached the level of the strongest classical programs.
Document type :
Conference papers
Complete list of metadatas

Cited literature [13 references]  Display  Hide  Download

https://hal.inria.fr/inria-00149859
Contributor : Rémi Coulom <>
Submitted on : Tuesday, May 29, 2007 - 10:44:26 AM
Last modification on : Thursday, February 21, 2019 - 10:52:49 AM
Long-term archiving on : Thursday, April 8, 2010 - 6:06:29 PM

File

MMGoPatterns.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : inria-00149859, version 1

Collections

Citation

Rémi Coulom. Computing Elo Ratings of Move Patterns in the Game of Go. Computer Games Workshop, Jun 2007, Amsterdam, Netherlands. ⟨inria-00149859⟩

Share

Metrics

Record views

1765

Files downloads

2048