论文信息 - On the Parallelization of UCT

On the Parallelization of UCT

We present three parallel algorithms for UCT. For 9 ×9 Go, they all improve the results of the programs that use them against G NU O 3.6. The simplest one, the single-run algorithm, uses very few commu nications and shows improvements comparable to the more complex ones. Further i mprovements may be possible sharing more information in the multiple-runs a lgorithm.

T. Cazenave | Nicolas Jouandeau

[1] Bernd Brügmann Max-Planck. Monte Carlo Go , 1993 .

[2] Greg Burns,et al. LAM: An Open Cluster Environment for MPI , 2002 .

[3] Sylvain Gelly,et al. Exploration exploitation in Go: UCT for Monte-Carlo Go , 2006, NIPS 2006.

[4] Olivier Teytaud,et al. Modification of UCT with Patterns in Monte-Carlo Go , 2006 .

[5] Rémi Coulom,et al. Efficient Selectivity and Backup Operators in Monte-Carlo Tree Search , 2006, Computers and Games.

[6] Csaba Szepesvári,et al. Bandit Based Monte-Carlo Planning , 2006, ECML.

[7] Sylvain Gelly,et al. Modifications of UCT and sequence-like simulations for Monte-Carlo Go , 2007, 2007 IEEE Symposium on Computational Intelligence and Games.