A Model for Comparing Game Theory and Artificial Intelligence Decision Making Processes
Abstract
The purpose of this thesis was to develop an analytical tool to compare the use of an artificial intelligence decision making method to a game theory decision making method. The game theory method used was the two-player zero-sum gaming technique. The artificial intelligence method used was the minimax technique with an alpha-beta cutoff heuristic. The simulation program was written in the C programming language. The artificial intelligence method made calls to a forward chaining shell called C Language Production System, CLIPS. The original game theory method was developed by Capt Robert Palmer. It was written in Turbo Pascal and required some modification before use in this study. The program was converted to the C language, modified to fit its intended use, and fit with the artificial intelligence method described above. The program played reconnaissance satellites against ground mobile Strategic Relocatable Targets (SRTs). It implemented the methods described within this document to allow for a comparison of the two techniques. For an initial comparison, the measure of effectiveness was the number of times the satellite incorrectly predicted the SRT's movement.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1989
- Accession Number
- ADA215418
Entities
People
- Paul R. Andre
Organizations
- Air Force Institute of Technology