Top Reinforcement Learning Algorithms

One of the biggest driving forces in the recent advancements in AI has been reinforcement learning (RL). A simple definition of reinforcement learning is to train the machine to act as good as possible by giving it feedback for its action, which implies finding a policy that maximises the expected return. RL algorithms adopt slightly different approaches to train the agent for its actions. ​​However, to find the optimal policy and value, most of the RL algorithms follow a similar pattern with either a model-free or a model-based approach along with an on-policy or off-policy approach.  All RL algorithms have common terms that need to be understood before diving into the algorithms. Action (A): Moves that the agent makes. State (S): Current situation in the environment.