Dynamic programming
DeÖnition
(^1) Sis called the state space.
(^2) Ais called the action space.
(^3) rt(s,a)is the reward function at time periodt.
(^4) ft(s,a)is the transition function at time periodt.
(^5) Φt(s)! 2 Ais the correspondence of feasible actions at time period
t.