(Reinforcement Learning): In reinforcement learning, the episode is the recording of actions and states that an agent performed from a start state to an end state. Using the concept of “trial and error” for reinforcement learning, the episode is a specific series of trials and errors. For example, if an agent is trying to find a solution to a maze, the episode would be the series of movements the agent made to reach a particular outcome without having to backtrack or end at an obstacle.
Glossary of Artificial Intelligence (AI), Machine Learning (ML), and Big Data Terms
Episode
