General Video Game AI Controller - Integrating Three Algorithms to Bring a New Solution

Abstract

This paper presents the framework, rules, games, controllers, and results of the General Video Game Playing Competition. The competition proposes the challenge of creating controllers of an artificial intelligence for general video game play, where a single agent must be able to play many different and wide games, some of them unknown games as well as in principle unlimited range of games. This work describes the implementation of the controller by using A-Star Pathfinding, Monte Carlo and Genetic Algorithm. We also use combinations of the above mentioned algorithm to achieve better results. The paper also analyzes the performance of the different approaches submitted, and finally proposes future tracks for the competition.