Date of Award

5-2024

Degree Type

Thesis

Degree Name

M.S.

Degree Program

Computer Science

Department

Computer Science

Major Professor

Dr. Ben Samuel

Second Advisor

Dr. Atriya Sen

Third Advisor

Dr. Abdullah Al Redwan Newaz

Fourth Advisor

Md Tamjidul Hoque

Abstract

This paper describes the creation and development of an implementation of the NeuroEvolution of Augmenting Topologies (NEAT) architecture to train an agent to play Super Mario Brothers. Building off of a basic implementation of NEAT, this thesis project shows the process of refining the fitness calculation that ranks the networks in the population and also defines the creation and application of a dataset to train the agent. The use of a dataset to train an agent is a novel idea in the world of reinforcement learning because, generally, reinforcement learning trains an agent to complete a singular task like the pole balancing problem. Training an agent to play something as complex as a video game, however, requires that an agent is exposed to as many different situations that occur within the game as possible. The goal of this thesis project is to create an agent that has a robust general understanding of how to play the game, such that it is able to react to new situations that were not seen in training. The results of this thesis project show that this generalized understanding is possible via neuroevolution, when given enough training time, a properly designed fitness calculation, and a properly applied dataset of scenarios.

Rights

The University of New Orleans and its agents retain the non-exclusive license to archive and make accessible this dissertation or thesis in whole or in part in all forms of media, now or hereafter known. The author retains all other ownership rights to the copyright of the thesis or dissertation.

Download

Share

COinS