ByteFight Game Agent

Developer

Used a lightweight sandbox environment to experiment with simple game agents using reinforcement learning. The project focused on play a zero-sum game against an opponent agent in a grid-like environment. I implemented reward-shaping strategies, state representations, and policy exploration techniques to compare agent performance. Our agent placed 5th out of 100 teams in the compeititon.

Conducted as a project for an Intro to Artificial Intelligence (CS3600) at The Georgia Institute of Technology