This is a Machine Learning project made by myself, containing several algorithms from the Reinforcement Learning world (DQN, A2C, PPO) with Single and Multi-Agent approaches, to try and minimize waiting time for cars in the SUMO traffic simulator.