Projektarbeit - Flatland Challenge Project work about train guidance using model-free multi-agent reinforcement learning by Ralph Meier & Dano Roost. Arxiv link to associated paper: Link PDF of thesis: PDF Download