Bruno Bouzy: Associating Shallow and Selective Global Tree Search with Monte Carlo for 9*9 Go. Computers and Games Bruno Bouzy of Paris Descartes, CPSC, Paris (Paris 5) with expertise in: Artificial Intelligence. Read 73 publications, and contact Bruno Bouzy on ResearchGate. Bruno Bouzy is a player and programmer from France. Born in , his highest rank was 3 dan. He was vice champion of France, losing in the.

Author: Gardasar Akinris
Country: Finland
Language: English (Spanish)
Genre: Education
Published (Last): 4 September 2010
Pages: 102
PDF File Size: 7.29 Mb
ePub File Size: 15.62 Mb
ISBN: 339-3-86713-596-9
Downloads: 50363
Price: Free* [*Free Regsitration Required]
Uploader: Mikagis

However, it does have some drawbacks because bpuzy evaluation of a move from a random game in which it was played at a late stage is less reliable than when it is played at an early stage. The brunk domain-dependent knowledge required is the definition of an eye. To date, we have estimated the value of a gouzy by considering only the final scores of the random games where it had been played. In figure 3 the point C is good for white and bad for black.

The former is the basic idea, the latter is what was performed in Gobble. A review of game-tree pruning. We believe that such an approach does not have the drawback of go global tree search with very little domain-dependent knowledge no termination and the drawback of domain-dependent move generation no verification.

To support our view, section 2 describes the related work about Monte Carlo applied to go. The time to carry out these tests is proportional to the time spent to play one random game.

The game of Go is one of the games that still withstand classical Artificial Intelligence approaches. The maximal threshold is fixed to 10, multiplied by the number of legal moves. However, this can turn out to be a fairly dangerous trick. Static eye in “the many faces of go”.

Fur- thermore, simulated annealing was used to control the probability that a move could be played out of order. Let us start the random games from the root by two given moves, one move for the friendly side, and, then, one move for the opponent, and make statistics on the terminal position evaluation for each node situated at depth 2 in the min-max tree.


The problems related to Computer Go require new AI problem nruno methods. The main idea was to use a very little domain-dependent knowledge approach. Nevertheless, this approach looks promising.

Prospective methods of programming the game of Go will probably be of interest in other domains as well.

In incomplete information games, such as poker [Billings et al. This is important for the random program not to play a move in an eye. Therefore, this result between two very different architectures is quite enlightening. Artificial Intelligencepages — The minimal number of random games without pruning brruno set to To evaluate a position, play a given number bouzh completely random games to the end – without filling the eyes – and score them.

On one hand, mathematical morphology is a very powerful tool within image processing community. Its move decision process is described in [Bouzy, ]. Information Sciences, pages — Besides, we have tried to enhance our programs with a depth-two tree search which did not work well. How do the uses of transpositions and progressive pruning compare in strength?

It is easy to add a simple tactical module which reads ladder. Create your web page Haltools: Vegos is a recent go program available on the web [Kaminski, ].

Bruno Bouzy

We can conclude that 10, random games per move is a good compromise when using transpositions. At depth-one nodes, the value is computed by using the min rule. Bruno Bouzy 1 AuthorId: Here are the results of 40 9×9 games and 20 13×13 games between PP and Indigo Because strings, liberties and intersection accessibilities are updated incre- mentally during the random games, the number of moves per second is almost constant and the time to play a game is proportional to the board size.

Progressive pruning does not need transpositions, temperature or simulated annealing. Oleg uses the basic data structure of GnuGo that is already very well optimized by the GnuGo team [Bump, ]. Both the minimal number of random games and the maximal threshold remain constant and 10, respectively. In this paper, we investigate the use of contextual knowledge in order to simpli fy knowledge representation in very complex domains and systems. Since the precision of the expected value depends on the square of the number of random games, there is no need to gain 20 percent in speed, which would only bring about a 10 percent improvement in the precision.


If Go dealt only with connections and not with captures, then perhaps it might be called Hex, and this problem would not arise. Previous works assessed that Qlearning surpassed Nash-based … More.

In these two cases, the move with the highest expected outcome is chosen. In order to choose a move in a given position, Gobble played a large number of almost random games from this position to the end, and scored them.

In this context, each module is independent of the other one, and does not use the strength of the other one. Probably the best moves are played early and thus, get a more accurate evaluation. Abstract Since the beginning of AI, mind games have been studied as relevant application fields. It has already been considered theoretically within the framework of [Rivest, ].

In this respect, this approach fills the gap left by global tree search in computer go no termination and left by move generation no verification.

Bruno Bouzy at Sensei’s Library

Second, we must play more of them to have an accurate evaluation of the moves we did it with This phenomenon happens when captures have already occurred at the time when the move is played. Simulated annealing normally has an evaluation that depends only on the current state in the case of Gobble, a state is the lists of moves for both players ; instead in Gobble the evaluation of a state is the average of all the random games that are based on all the states reached so far.

On a 2 Ghz computer, Olga plays 7, random 9×9 games per second and Oleg 10, This approach is based on statistics or Monte Carlo methods.