GAIG Game AI Research Group @ QMUL

Toward General Mathematical Game Playing Agents


Abstract

General game playing AI and general video game playing AI are both active research areas. Mathematical games, like prisoner’s dilemma, rock-paper-scissors, or the snowdrift game can also be played by general purpose agents. An agent representation in which agents play knowing only their own score and their opponent’s score in a previous round is used to enable general mathematical game playing. Agents are trained to recognize when the game being played has changed and trained to play competently on diverse sets of games. This domain, simpler than general game playing or general video game playing, is offered as a test lab for concepts in the creation of general AI. It is demonstrated that agents can be trained to play three different coordination games competently. The agents also learn to play prisoner’s dilemma and a coordination game that reverses which moves lead to high payoffs.

Cite this work

@inproceedings{ashlock2018toward,
author= {Ashlock, Daniel and Kim, Eun-Youn and Perez-Lebana, Diego},
title= {{Toward General Mathematical Game Playing Agents}},
year= {2018},
booktitle= {{Proc. of the IEEE Conference on Computational Intelligence and Games (CIG)}},
month= {Aug},
pages= {1--7},
abstract= {General game playing AI and general video game playing AI are both active research areas. Mathematical games, like prisoner’s dilemma, rock-paper-scissors, or the snowdrift game can also be played by general purpose agents. An agent representation in which agents play knowing only their own score and their opponent’s score in a previous round is used to enable general mathematical game playing. Agents are trained to recognize when the game being played has changed and trained to play competently on diverse sets of games. This domain, simpler than general game playing or general video game playing, is offered as a test lab for concepts in the creation of general AI. It is demonstrated that agents can be trained to play three different coordination games competently. The agents also learn to play prisoner’s dilemma and a coordination game that reverses which moves lead to high payoffs.},
}

Comments

Content