Proposal Title : Application of Evolutionary Computation to Opponent Modeling in Games of Imperfect

The video game industry is an area of high and increasing profitability, with $US 6 billion spent by U.S. consumers alone in the year 2001 [1]. In order to attract buyers, there is increasing demand to simulate human opponents artificially within games. As of present, artificial opponents commonly have their game-playing strategies coded in advance, often resulting in overly simplistic and predictable opponents whose flaws are easily exploitable. Games such as Bridge, Poker and Scrabble are games of imperfect information, meaning that not all the information about the state of the game is known (e.g. hidden opponent cards in Poker). Due to the non-deterministic nature of such games and the need for opponent modeling, the task of determining satisfactory strategies is extremely broad and difficult to be programmed for in advance. To make games more interactive, it is necessary to have computer opponents that learn and adapt; opponents who are free to function within the rules of the game, rather than some predefined subset conceived by the programmer. The ideal situation would be to create an artificial opponent that is able to adopt various game-playing strategies and appear indistinguishable from a human player.