Code repository for the research project "You Play Ball, I Play Ball: Bayesian Multi-Agent Reinforcement Learning for Slime Volleyball".
Presented at National University of Singapore - 17th School of Computing Term Project Showcase (17th STePS).
Demo & Video
In Slime Volleyball, a two-player competitive game, we investigate how improves AI players’ learning in 3 ways: 1) , 2) and 3) , in the domain of multi-agent reinforcement learning (MARL).
We show that by modelling uncertainty, Bayesian methods improve MARL training in 4 ways: 1) , 2) , 3) and 4) , and through experiments using TensorFlow Probability and Stable Baselines, we present interesting differences in agent behaviour.
We contribute code for 3 functionalities: 1) integrated into Stable Baselines, 2) for Stable Baselines (previously with only single-agent support) and 3) for Slime Volleyball Gym.