Reinforcement Learning: Type of machine learning, where a computer agent learns to perform a task through a process of repeated trial and error within a dynamic environment. The goal is to learn what to do, how to map situations to take certain actions that minimize a numerical reward signal. Eg. RL trained computer gains can achieve greater levels than world’s best human beings.
In a control system, we need to determine correct inputs which are actions into the system to generate desired system behaviour. Thus, it decides which action to choose that results in the desired behaviour by the system.
Reinforcement learning is one the main three machine learning techniques, where the other two are supervised learning and unsupervised learning.
The goal of reinforcement learning is not to cluster or label but to find the best sequence of actions for optimal outcome.
Anatomy of RL: Policy takes state observations to map them to the action. The goal of the agent is to use reinforcement learning algorithm to learn the best policy and thus given any states take the optimal action which results in the best reward in the long run.
Other important topics:
