Reinforcement Learning: A subfield of AI that deals with the development of models and algorithms that can learn from trial and error, it can be used to train agents to make decisions and to control systems.