TY - GEN
T1 - GA-gammon
T2 - 5th Mexican International Conference on Artificial Intelligence, MICAI 2006
AU - Irineo-Fuentes, Oscar
AU - Cruz-Cortés, Nareli
AU - Rodríguez-Henríquez, Francisco
AU - Ortiz-Arroyo, Daniel
AU - Larsen, Henrik Legind
PY - 2006
Y1 - 2006
N2 - In this paper we describe a genetic algorithm approach able to confection strong backgammon automata players. We first prepared an initial vector of weights representing a set of heuristic strategies suggested by expert human players. Then, employing a genetic algorithm approach we were able to fine tune such initial vector of weights by repeatedly testing it against Pubeval, a strong benchmark player program. The vector of weights was therefore used as an evaluation function for performing a genetic heuristic selection of the best board positions during a game. Best GA-Gammon individuals so obtained were tested in separated 5000-game tournaments against Pubeval itself, and Fuzzeval, a fuzzy controllerbased player. Our experimental results indicate that the best individuals generated by GA-Gammon show similar performance than Pubeval. Furthermore, GA-Gammon consistently outperforms Fuzzeval.
AB - In this paper we describe a genetic algorithm approach able to confection strong backgammon automata players. We first prepared an initial vector of weights representing a set of heuristic strategies suggested by expert human players. Then, employing a genetic algorithm approach we were able to fine tune such initial vector of weights by repeatedly testing it against Pubeval, a strong benchmark player program. The vector of weights was therefore used as an evaluation function for performing a genetic heuristic selection of the best board positions during a game. Best GA-Gammon individuals so obtained were tested in separated 5000-game tournaments against Pubeval itself, and Fuzzeval, a fuzzy controllerbased player. Our experimental results indicate that the best individuals generated by GA-Gammon show similar performance than Pubeval. Furthermore, GA-Gammon consistently outperforms Fuzzeval.
UR - http://www.scopus.com/inward/record.url?scp=34547660220&partnerID=8YFLogxK
U2 - 10.1109/MICAI.2006.23
DO - 10.1109/MICAI.2006.23
M3 - Contribución a la conferencia
AN - SCOPUS:34547660220
SN - 0769527221
SN - 9780769527222
T3 - Proceedings - Fifth Mexican International Conference on Artificial Intelligence, MICAI 2006
SP - 207
EP - 216
BT - Proceedings - Fifth Mexican International Conference on Artificial Intelligence, MICAI 2006
Y2 - 13 November 2006 through 17 November 2006
ER -