The existence of good Markov strategies for decision processes with general payoffs

Theodore P. Hill, Victor Pestien

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

For countable-state decision processes (dynamic programming problems), a general class of objective functions is identified for which it is shown that good Markov strategies always exist. This class includes product and lim inf rewards, as well as practically all the classical dynamic programming expected payoff functions.

Original languageEnglish (US)
Pages (from-to)61-76
Number of pages16
JournalStochastic Processes and their Applications
Volume24
Issue number1
DOIs
StatePublished - Jan 1 1987

Fingerprint

Dynamic programming
Dynamic Programming
Reward
Countable
Objective function
Strategy
Class

Keywords

  • dynamic programming
  • Markov decision process
  • Markov strategy

ASJC Scopus subject areas

  • Statistics and Probability
  • Modeling and Simulation
  • Applied Mathematics

Cite this

The existence of good Markov strategies for decision processes with general payoffs. / Hill, Theodore P.; Pestien, Victor.

In: Stochastic Processes and their Applications, Vol. 24, No. 1, 01.01.1987, p. 61-76.

Research output: Contribution to journalArticle

@article{da6ecad4bf2d452291a1bd2444ef4f6a,
title = "The existence of good Markov strategies for decision processes with general payoffs",
abstract = "For countable-state decision processes (dynamic programming problems), a general class of objective functions is identified for which it is shown that good Markov strategies always exist. This class includes product and lim inf rewards, as well as practically all the classical dynamic programming expected payoff functions.",
keywords = "dynamic programming, Markov decision process, Markov strategy",
author = "Hill, {Theodore P.} and Victor Pestien",
year = "1987",
month = "1",
day = "1",
doi = "10.1016/0304-4149(87)90028-7",
language = "English (US)",
volume = "24",
pages = "61--76",
journal = "Stochastic Processes and their Applications",
issn = "0304-4149",
publisher = "Elsevier",
number = "1",

}

TY - JOUR

T1 - The existence of good Markov strategies for decision processes with general payoffs

AU - Hill, Theodore P.

AU - Pestien, Victor

PY - 1987/1/1

Y1 - 1987/1/1

N2 - For countable-state decision processes (dynamic programming problems), a general class of objective functions is identified for which it is shown that good Markov strategies always exist. This class includes product and lim inf rewards, as well as practically all the classical dynamic programming expected payoff functions.

AB - For countable-state decision processes (dynamic programming problems), a general class of objective functions is identified for which it is shown that good Markov strategies always exist. This class includes product and lim inf rewards, as well as practically all the classical dynamic programming expected payoff functions.

KW - dynamic programming

KW - Markov decision process

KW - Markov strategy

UR - http://www.scopus.com/inward/record.url?scp=38249036444&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=38249036444&partnerID=8YFLogxK

U2 - 10.1016/0304-4149(87)90028-7

DO - 10.1016/0304-4149(87)90028-7

M3 - Article

AN - SCOPUS:38249036444

VL - 24

SP - 61

EP - 76

JO - Stochastic Processes and their Applications

JF - Stochastic Processes and their Applications

SN - 0304-4149

IS - 1

ER -