Automatically acquiring domain knowledge for adaptive game AI using evolutionary learning

  • Authors:

  • Marc J. V. Ponsen;Héctor Muñoz-Avila;Pieter Spronck;David W. Aha

  • Affiliations:

  • Dept. of Computer Science & Engineering, Lehigh University, Bethlehem, PA;Dept. of Computer Science & Engineering, Lehigh University, Bethlehem, PA;IKAT, Maastricht University, Maastricht, MD, The Netherlands;Navy Center for Applied Research in AI, Naval Research Laboratory, Washington, DC

  • Venue:
  • IAAI'05 Proceedings of the 17th conference on Innovative applications of artificial intelligence - Volume 3
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

Game AI is the decision-making process of computer-controlled opponents in computer games. Adaptive game AI can improve the entertainment value of computer games. It allows computer-controlled opponents to automatically fix weaknesses in the game AI and respond to changes in human-player tactics. Dynamic scripting is a recently developed approach for adaptive game AI that learns which tactics (i.e., action sequences) an opponent should select to play effectively against the human player. In previous work, these tactics were manually generated. We introduce AKADS; it uses an evolutionary algorithm to automatically generate such tactics. Our experiments show that it improves dynamic scripting's performance on a real-time strategy (RTS) game. Therefore, we conclude that high-quality domain knowledge (i.e., tactics) can be automatically generated for strong adaptive AI opponents in RTS games. This reduces the time and effort required by game developers to create intelligent game AI, thus freeing them to focus on other important topics (e.g., storytelling, graphics).