Markov Decision Process

Markov Decision Process Framework

AI, ML
AI, ML, Reinforcement Learning, Markov Decision Process

Markov Decision Process Framework #

A Markov Decision Process (MDP) is a mathematical framework for modelling sequential decisions. It describes the situations an agent may encounter, the actions it may take, how the environment may change, and the rewards produced by those changes.

Bandit problems ask which action is best in a single recurring situation. An MDP adds changing states: an action affects not only the immediate reward but also the situation faced next.