Too Long; Didn't Read
Reinforcement Learning is a branch of machine learning focusing on learning with interaction. The ultimate goal of any learning problem in reinforcement setting is to achieve a certain goal. This goal can be to win a game of chess or balancing a pole. With more and more interactions, the uncertainty of the environment is reduced. Agent-Environment interaction is a cycle of agent sensing the environment, taking an appropriate action and receiving reward from environment for its action. An action which brings the state of environment to a more acceptable state gives higher reward.