Development of Artificial Intelligence for Poker Game

OData support
Supervisor:
Dr. Juhász Sándor
Department of Automation and Applied Informatics

Abstract

The topic of my essay is the nowadays increasingly popular card game – poker. The popularity of the poker game is due to its supposed simplicity, as the rules and the course of the game are easy to acquire. But on closer examination we have to realize that from the calculation discipline respect this game is very complicated and complex.[1]

Thus, when we try to automate the game, and write a program simulating human decisions, we face the fact that the result of a sudden and logical decision of a human can be very difficult to a program and its programmer.

In this essay we will deal with the creation of a „poker robot” that imitates the human decisions in the game. There are a lot of books [3] dealing with different poker strategies, from which we try to make a good artificial intelligence by finding the right parameters for the algorithms [2].

The decision-making algorithms will be tested by running simulations within a framework. We seek the optimal parameters for the determination of the weight parameters of the decision-making algorithm by the evaluation of the results of the simulations.

In the end of the essay we will demonstrate a working poker robot that will cope with other robots during the simulations while keeping the unpredictable property of a human player. We will also examine how to expand the test of our robot and how to upgrade it’s results against computers and human players as well.

Downloads

Please sign in to download the files of this thesis.