Reinforcement Learning Agent Learns to play Checkers

Checkers is a well-known board game. Using machine learning to develop an algorithm that can learn games like Checkers, Chess, etc has always been a fascination for man. Since the development of modern computers there have been numerous algorithms that are capable of performing well against humans. However most algorithms are artificial intelligence based, in other words the algorithm mimics human intelligence without ac- tually learning the game. In this project, we will try applying reinforcement learning to develop an agent that will learn to play checkers. We will test its capability by letting the agent play an opponent which makes either random moves or moves selected by an AI.