Deep reinforcement learning. In scikit-learn. In less than 50 effective lines. - View it on GitHub