10 Reinforcement learning

What is reinforcement learning?

Reinforcement learning is an area of machine learning inspired by behavioral psychology. The concept of reinforcement learning is based on cumulative rewards or penalties for the actions an agent takes in a dynamic environment. Think about a young dog growing up. The dog is the agent in an environment that is our home. When we want the dog to sit, we might simply say, “Sit.” The dog doesn’t understand English, so we might nudge it by lightly pushing down on its back. After the dog sits, we pet it or give it a treat—a welcome reward. We need to repeat this process many times, but eventually, we positively reinforce the idea of sitting for the dog. The trigger in the environment is saying “Sit”; the behavior learned is sitting; the reward is petting or treats.

The inspiration for reinforcement learning

Problems that reinforcement learning can solve

The life cycle of reinforcement learning

Simulation and data: The agent’s environment

Training with the simulation using Q-learning

Testing with the simulation and Q-table

Measuring the performance of training

Model-free and model-based learning

Deep learning approaches to reinforcement learning

Training with an ANN

Use cases for reinforcement learning