Associating domain-dependent knowledge and Monte Carlo approaches within a Go program

Bouzy, Bruno

doi:10.1016/j.ins.2004.04.010

Cited by 47 publications

(46 citation statements)

References 12 publications

(18 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Using some simple 3 × 3 patterns inspired by Indigo [11] (similar patterns can also be found in [12]), our random simulation is likely to have more meaningful sequences in random simulations than before, which has improved significantly the level of MoGo. Essentially, we use patterns to create meaningful sequences in simulations by finding local answers.…”

Section: B Improving Simulation With Domain-dependent Knowledgementioning

confidence: 99%

Modifications of UCT and sequence-like simulations for Monte-Carlo Go

Wang

Gelly

2007

2007 IEEE Symposium on Computational Intelligence and Games

View full text Add to dashboard Cite

Abstract-Algorithm UCB1 for multi-armed bandit problem has already been extended to Algorithm UCT which works for minimax tree search. We have developed a Monte-Carlo program, MoGo, which is the first computer Go program using UCT. We explain our modification of UCT for Go application and also the sequence-like 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 Computer-Go program on 9 × 9 Go board.

show abstract

Section: B Improving Simulation With Domain-dependent Knowledgementioning

confidence: 99%

Modifications of UCT and sequence-like simulations for Monte-Carlo Go

Wang

Gelly

2007

2007 IEEE Symposium on Computational Intelligence and Games

View full text Add to dashboard Cite

show abstract

“…However, if Black plays well, C and F never get connected, and the value of the threat is much lower. A possible solution to the problem of the overevaluation of threats could be to make the program play better during the Monte-Carlo games [Bouzy 2005]. …”

Section: Resultsmentioning

confidence: 99%

Goal threats, temperature, and Monte-Carlo Go

Cazenave

2009

Games of No Chance 3

View full text Add to dashboard Cite

ABSTRACT. Keeping the initiative, i.e., playing sente moves, is important in the game of Go. This paper presents a search algorithm for verifying that reaching a goal is sente on another goal. It also presents how goals are evaluated. The evaluations of the goals are based on statistics performed on almost random games. Related goals, such as goals and associated threatened goals, are linked together to form simple subgames. An approximation of the temperature is computed for each move that plays in a simple subgame. The move with the highest temperature is chosen. Experimental results show that using the method improves a Go program.

show abstract

“…In that sense, designing and calibrating a MC engine remains an open challenge: one has to intensively experiment a modification in order to validate it. Various shapes are defined in [2,13,12]. [13] uses patterns and expertise as explained in Algorithm 1.…”

Section: Improving Monte-carlo (Mc) Simulationsmentioning

confidence: 99%

Adding Expert Knowledge and Exploration in Monte-Carlo Tree Search

Chaslot

Fiter

Hoock

et al. 2010

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. We present a new exploration term, more efficient than classical UCT-like exploration terms and combining efficiently expert rules, patterns extracted from datasets, All-Moves-As-First values and classical online values. As this improved bandit formula does not solve several important situations (semeais, nakade) in computer Go, we present three other important improvements which are central in the recent progress of our program MoGo:-We show an expert-based improvement of Monte-Carlo simulations for nakade situations; we also emphasize some limitations of this modification. -We show a technique which preserves diversity in the Monte-Carlo simulation, which greatly improves the results in 19x19. -Whereas the UCB-based exploration term is not efficient in MoGo, we show a new exploration term which is highly efficient in MoGo. MoGo recently won a game with handicap 7 against a 9Dan Pro player, Zhou JunXun, winner of the LG Cup 2007, and a game with handicap 6 against a 1Dan pro player, Li-Chen Chien. 1

show abstract

Associating domain-dependent knowledge and Monte Carlo approaches within a Go program

Cited by 47 publications

References 12 publications

Modifications of UCT and sequence-like simulations for Monte-Carlo Go

Modifications of UCT and sequence-like simulations for Monte-Carlo Go

Goal threats, temperature, and Monte-Carlo Go

Adding Expert Knowledge and Exploration in Monte-Carlo Tree Search

Contact Info

Product

Resources

About