Optimal policies for constrained average-cost Markov decision processes

We give mild conditions for the existence of optimal solutions for a Markov decision problem with average cost, under m constraints of the same kind, in Borel actions and states spaces. Moreover, there is an optimal policy that is a convex combination of at most m + 1 deterministic policies.

Bibliographic Details
Main Authors: González Hernández, Juan, Villarreal, César Emilio
Format: Article
Language:English
Published: 2009
Subjects:
Online Access:http://eprints.uanl.mx/1957/1/Top.pdf
Description
Summary:We give mild conditions for the existence of optimal solutions for a Markov decision problem with average cost, under m constraints of the same kind, in Borel actions and states spaces. Moreover, there is an optimal policy that is a convex combination of at most m + 1 deterministic policies.