inria-00117266, version 2
Modification of UCT with Patterns in Monte-Carlo Go
N° RR-6062 (2006)
Abstract: Algorithm UCB1 for multi-armed bandit problem has already been extended to Algorithm UCT (Upper bound Confidence for Tree) which works for minimax tree search. We have developed a Monte-Carlo Go program, MoGo, which is the first computer Go program using UCT. We explain our modification of UCT for Go application and also the intelligent random simulation with patterns which has improved significantly the performance of MoGo. UCT combined with pruning techniques for large Go board is discussed, as well as parallelization of UCT. MoGo is now a top level Go program on $9\times9$ and $13\times13$ Go boards.
- 1:
- INRIA – CNRS : UMR8623 – Université Paris XI - Paris Sud
- 2:
- CNRS : UMR7641 – Université de Versailles Saint-Quentin-en-Yvelines – Polytechnique - X
- 3:
- INRIA – CNRS : UMR8146 – Université Lille I - Sciences et technologies – Université Lille III - Sciences humaines et sociales – Ecole Centrale de Lille
- Domain : Computer Science/Learning
Computer Science/Artificial Intelligence - Internal note : RR-6062
- Available versions : v1 (2006-11-30) v2 (2006-12-12) v3 (2006-12-21)
- inria-00117266, version 2
- http://hal.inria.fr/inria-00117266
- oai:hal.inria.fr:inria-00117266
- From:
- Submitted on: Tuesday, 12 December 2006 10:35:11
- Updated on: Tuesday, 12 December 2006 10:35:24





Associated documents
Export