Finite State Markovian Decision Processes |
Contents
Computational Example | 15 |
Some Existence Theorems | 19 |
Expected Average Cost Problem | 25 |
Copyright | |
15 other sections not shown
Other editions - View all
Common terms and phrases
A₁ action Appendix approach arbitrary assume average cost called Chapter compact Computational condition consider constraints continuous convex Corollary cost cost criterion defined denote Derman determine discounted cost dual problem dynamic programming equality equations example exists fact feasible finite follows formulation function given Hence holds initial inspection K₁ least Lemma limit linear programming problem Math maximize method of successive minimize object observed obtained optimal policy optimal solution periodic policy improvement procedure possible primal problem probability Proof prove R₁ R₂ random recurrent Remarks replace respect satisfies selected sequence solution solving space stochastic stopping successive approximations Suppose taken takes action Theorem transient uniquely unit v₁ variables VR(i Y₁ YR(i ΣΣ