In this chapter, we will learn about reinforcement learning. As the name suggests, with this method, optimal strategies are discovered through reinforcing or rewarding certain behavior and penalizing other behavior. The basic idea for this type of machine learning is to use an agent that performs actions towards a goal in an environment. We will explore this machine learning technique by using the ReinforcementLearning package in R to compute a policy for the agent to win a game of tic-tac-toe.
While this may seem like a simple game, it is a good environment for investigating reinforcement learning. We will learn how to structure input data for reinforcement learning, which is the same format for tic-tac-toe as for more complex games. We will learn how to compute a policy using the input data to provide the agent with the optimal strategy...
