Here is my python source code for training an agent to play Tetris. It could be seen as a very basic example of Reinforcement Learning's application.