We apply genetic programming to the evolution of
strategies for playing the game of backgammon. We
explore two different strategies of learning: using a
fixed external opponent as teacher, and letting the
individuals play against each other. We conclude that
the second approach is better and leads to excellent
results: Pitted in a 1000-game tournament against a
standard benchmark player Pubeval our best evolved
program wins 62.4 percent of the games, the highest
result to date. Moreover, several other evolved
programs attain win percentages not far behind the
champion, evidencing the repeatability of our
approach.
%0 Journal Article
%1 azaria:2005:GPEM
%A Azaria, Yaniv
%A Sipper, Moshe
%D 2005
%J Genetic Programming and Evolvable Machines
%K STGP, algorithms, backgammon, coevolution demes, genetic programming, self-learning,
%N 3
%P 283--300
%R doi:10.1007/s10710-005-2990-0
%T GP-Gammon: Genetically Programming Backgammon
Players
%U http://www.cs.bgu.ac.il/~sipper/papabs/gpgammon.pdf
%V 6
%X We apply genetic programming to the evolution of
strategies for playing the game of backgammon. We
explore two different strategies of learning: using a
fixed external opponent as teacher, and letting the
individuals play against each other. We conclude that
the second approach is better and leads to excellent
results: Pitted in a 1000-game tournament against a
standard benchmark player Pubeval our best evolved
program wins 62.4 percent of the games, the highest
result to date. Moreover, several other evolved
programs attain win percentages not far behind the
champion, evidencing the repeatability of our
approach.
@article{azaria:2005:GPEM,
abstract = {We apply genetic programming to the evolution of
strategies for playing the game of backgammon. We
explore two different strategies of learning: using a
fixed external opponent as teacher, and letting the
individuals play against each other. We conclude that
the second approach is better and leads to excellent
results: Pitted in a 1000-game tournament against a
standard benchmark player Pubeval our best evolved
program wins 62.4 percent of the games, the highest
result to date. Moreover, several other evolved
programs attain win percentages not far behind the
champion, evidencing the repeatability of our
approach.},
added-at = {2008-06-19T17:35:00.000+0200},
author = {Azaria, Yaniv and Sipper, Moshe},
biburl = {https://www.bibsonomy.org/bibtex/2c9847140f8ec9ab82ab41542318325cd/brazovayeye},
doi = {doi:10.1007/s10710-005-2990-0},
interhash = {6fa5d4b83b74fd39bc396c73a4683b57},
intrahash = {c9847140f8ec9ab82ab41542318325cd},
issn = {1389-2576},
journal = {Genetic Programming and Evolvable Machines},
keywords = {STGP, algorithms, backgammon, coevolution demes, genetic programming, self-learning,},
month = {September},
note = {Published online: 12 August 2005},
notes = {ECJ},
number = 3,
pages = {283--300},
timestamp = {2008-06-19T17:36:01.000+0200},
title = {{GP}-Gammon: Genetically Programming Backgammon
Players},
url = {http://www.cs.bgu.ac.il/~sipper/papabs/gpgammon.pdf},
volume = 6,
year = 2005
}