Ben Rayment

Honours, 2020

For my research this year, I was interested in finding a solution to the question, what reinforcement learning algorithms are available and applicable to generate realistic game simulations? I investigated this through designing, experimenting, and evaluating a soccer simulation between two teams of four players using multiple reinforcement learning algorithms.

These algorithms allow developers to create agents that can think and train for themselves without explicitly programming them to carry out tasks. I’m carrying out this research not just for myself but to show the world that reinforcement learning is not just for games industry but for the entire world.

Ben Rayment

Selected Portfolio

Soccer Game Simulation

Soccer Game Simulation