Abstract

Poker provides an excellent testbed for studying decision-making under conditions of uncertainty. There are many benefits to be gained from designing and experimenting with poker programs. It is a game of imperfect knowledge, where multiple competing agents must understand estimation, prediction, risk management, deception, counterdeception, and agent modeling. New evaluation techniques for estimating the strength and potential of a poker hand are presented. This thesis describes the implementation of a program that successfully handles all aspects of the game, and uses adaptive opponent modeling to improve performance. Acknowledgements The author would like to acknowledge the following: ffl A big thanks to Jonathan Schaeffer, for being an excellent supervisor. ffl Darse Billings, for all his time, ideas and insightful input. ffl Duane Szafron, for contributing to the poker research group. ffl The National Sciences and Engineering Research Council, for providing financial support...

Citations