Comparing recent assumptions for the existence of average optimal stationary policies

Authors:
Rolando Cavazos-Cadena;Linn I Sennott
Affiliations:
Universidad Autonoma Agraria Antonio Narro, Buenavista 25315, Saltillo, COAH, Mexico;Illinois State University, Normal, IL 61761, USA
Venue:
Operations Research Letters
Year:
1992

Citing 3
Cited 4

Control of Markov chains with long-run average cost criterion: the dynamic programming equations

SIAM Journal on Control and Optimization
Introduction to Stochastic Dynamic Programming: Probability and Mathematical

Introduction to Stochastic Dynamic Programming: Probability and Mathematical
A new condition for the existence of optimal stationary policies in average cost Markov decision processes

Operations Research Letters

Optimal control of a production-inventory system with customer impatience

Operations Research Letters
On strong average optimality of markov decision processes with unbounded costs

Operations Research Letters
The value iteration method for countable state Markov decision processes

Operations Research Letters
The convergence of value iteration in average cost Markov decision chains

Operations Research Letters

Quantified Score

Hi-index	0.00

Visualization

Abstract

We consider discrete time average cost Markov decision processes with countable state space and finite action sets. Conditions recently proposed by Borkar, Cavazos-Cadena, Weber and Stidham, and Sennott for the existence of an expected average cost optimal stationary policy are compared. The conclusion is that the Sennott conditions are the weakest. We also give an example for which the Sennott axioms hold but the others fail.