RL: Reinforcement learning (RL) is an area of machine learning inspired by behaviorist psychology[citation needed], concerned with how software agents ought to take actions in an environment so as to maximize some notion of cumulative reward.