02ReinforcementLearning1.5A, Bellman equation

views comments

The Bellman equation connects the Q-value Q(s,a) with the Q-values of state-action pairs Q(s',a') that are reachable in one step.

(updated March 21)