Deep Reinforcement Learning to train agents in a multiplayer First Person Shooter: some preliminary results | IEEE Conference Publication | IEEE Xplore