Strategic reasoning in complex domains: A comparative survey on scientific AI techniques to improve real-time strategy game AI.

Abstract

Over the past decade, scientific artificial intelligence research has picked up real-time strategy video games as testbed for research. The large amount of data, the dynamic environment and strategic depth of the game makes the genre a challenge for scientific artificial intelligence research. This thesis is a comparative survey of two artificial intelligence techniques from scientific research, Case-based reasoning and Dynamic Scripting, to create a strong real-time strategy game playing agent.