Programming a Texas Hold'em Poker AI

Programming an AI capable of playing No Limit Texas Hold’em Poker is a very demanding task since during game-play there is hidden information to deal with and random events can occur. These features are not present in games that were historically most studied by artificial intelligence (like chess or checkers), and so they pose a significant and novel challenge for AI research. In this book we guide the readers through the basics of Poker AI development by describing implementation details of the HoldemML framework. This framework can be viewed as support tool that allows users to create Poker AIs that combine the strategies of several human players, inferred from past games. In addition, any Poker player who stores records of his or her game logs can use them to create an AI whose strategy reassembles the player’s. We hope that the description of the HoldemML framework in this book will not only prove useful for interested programmers to create their own approaches to a Poker AI, but also further develop the Computer Poker research domain.