Project description
This project will focus on developing hyper-agent (i.e., portfolio or ensemble) techniques for general game playing. These hyper-agents do not provide any new game-playing strategies, but instead attempt to combine multiple existing AI agents together. This is often done by using machine learning techniques to identify the individual strengths/weaknesses of each AI in specific situations. Depending on the situation encountered, our hyper-agent model can utilise the combined knowledge of prior agents to make the best decision.
Assumed knowledge
Knowledge and experience working with basic machine learning techniques (such as that covered in COMP2712)
Note: You need to register interest in projects from different supervisors (not a number of projects with the one supervisor).
You must also contact each supervisor directly to discuss both the project details and your suitability to undertake the project.