Project III/IV (MATH3382/4072) 2018-19

How do computers play games?

Matthias Troffaes and Camila Caiado

In this project, students will learn about algorithms for optimizing game play in games that involve 1 or more players. We will start with an analysis of simple games such as tic-tac-toe or tower of Hanoi and learn how to identify viable and/or optimal strategies. We will then move to more complex games of the student's choosing and look at issues such as optimizing gameplay against specific opponents, dealing with large state spaces, and stress testing games for game design.

The aim of this project is to explore different methods to teach computers how to play games such as reinforcement learning, agent-based modelling, and other decision or operation research approaches.

This project requires students to have taken one or more of the following modules:

As pre- or co-requisites, the students will be taking or will have taken at least one of the following modules:

References

Sutton, R. S., Barto, A. G. (1998) Reinforcement Learning: An Introduction, Cambridge, MIT Press.
Kulkarni, P. (2012) Reinforcement and Systemic Machine Learning for Decision Making , Hoboken, Wiley & Sons.

email: Matthias Troffaes; Camila Caiado