Video Game Strategy Learning System using Reinforcement Learning and Inference Rules-Edición Única

Abstract

Machine Learning is synonymous of advanced computing. Machine Learning is a growing body of work that exists on the use of such techniques to solve real world problems. The complex and/or ill understood nature of many problem domains, such as data mining or process control, has led to the need for technologies which can adapt to the task they face. Machine Learning has been greeted with a certain amount of caution by video games developers, and until recently, has not been used in any major games releases. Many video game companies are currently looking at the possibility of making video games that can match the player’s ability by altering tactics and strategy, rather than by improving the ability of the opponents using a difficulty system. Even on the toughest difficulty settings of most video games (specially on First Person Shouter games like Doom, released in 1993 ), most players have a routine, which if successful, will mean that they win more often than not. However, they would surely not be so smug if the artificial intelligence (AI) could work out their favorite hiding places, or uncover their winning tactics and adapt to them. This could become a very important feature of future releases, as it would prolong game life considerably. The importance for a video game to learn depends on the video game itself. For some video games, the average player does not appreciate any significant advance if the game learns and the learning effort is a waste of time and money. Nevertheless, video games like Black & White (2001) have not been half as successful if it was not for the creatures that could be taught, through kindness or cruelty, to mimic their master.