reinforcement-learning

Definition

Value Function

A value function gives the expected return if you start in that state or state-action pair, and then act according to a particular policy forever after.