Deep Q-learning (DQN) for Multi-agent Reinforcement Learning (RL)