random search, hill climbing, policy gradient - View it on GitHub