MARVEL: Bringing Multi-Agent Reinforcement- Learning Based Variable Speed Limit Controllers Closer to Deployment | IEEE Journals & Magazine | IEEE Xplore