Reinforcement Learning in StarCraft II

The aim was to compare popular multi-agent reinforcement learning strategies with mono-agent strategies that make strategic decisions in a highly dynamic environment like StarCraft II.

The project is available on GitHub.