TY - JOUR
T1 - Progressive Strategies for Monte-Carlo Tree Search
AU - Chaslot, Guillaume M. J-b.
AU - Winands, Mark H. M.
AU - van den Herik, H. Jaap
AU - Uiterwijk, Jos W. H. M.
AU - Bouzy, Bruno
PY - 2008/11/1
Y1 - 2008/11/1
N2 - Monte-Carlo Tree Search (MCTS) is a new best-first search guided by the results of Monte-Carlo simulations. In this article, we introduce two progressive strategies for MCTS, called progressive bias and progressive unpruning. They enable the use of relatively time-expensive heuristic knowledge without speed reduction. Progressive bias directs the search according to heuristic knowledge. Progressive unpruning first reduces the branching factor, and then increases it gradually again. Experiments assess that the two progressive strategies significantly improve the level of our Go program Mango. Moreover, we see that the combination of both strategies performs even better on larger board sizes.
AB - Monte-Carlo Tree Search (MCTS) is a new best-first search guided by the results of Monte-Carlo simulations. In this article, we introduce two progressive strategies for MCTS, called progressive bias and progressive unpruning. They enable the use of relatively time-expensive heuristic knowledge without speed reduction. Progressive bias directs the search according to heuristic knowledge. Progressive unpruning first reduces the branching factor, and then increases it gradually again. Experiments assess that the two progressive strategies significantly improve the level of our Go program Mango. Moreover, we see that the combination of both strategies performs even better on larger board sizes.
U2 - 10.1142/S1793005708001094
DO - 10.1142/S1793005708001094
M3 - Article
SN - 1793-0057
VL - 4
SP - 343
EP - 357
JO - New Mathematics and Natural Computation
JF - New Mathematics and Natural Computation
IS - 3
ER -