Multiple-Environment Markov Decision Processes: Efficient
Analysis and Applications

D 2020

Multiple-Environment Markov Decision Processes: Efficient Analysis and Applications

CHATTERJEE, Krishnendu, Martin CHMELÍK, Deep KARKHANIS, Petr NOVOTNÝ, Amélie ROYER et. al.

Basic information

Original name

Multiple-Environment Markov Decision Processes: Efficient Analysis and Applications

Authors

CHATTERJEE, Krishnendu (356 India), Martin CHMELÍK (203 Czech Republic), Deep KARKHANIS (356 India), Petr NOVOTNÝ (203 Czech Republic, guarantor, belonging to the institution) and Amélie ROYER (250 France)

Edition

Palo Alto, Proceedings of the International Conference on Automated Planning and Scheduling, p. 48-56, 9 pp. 2020

Publisher

AAAI Press

Other information

Language

English

Type of outcome

Stať ve sborníku

Field of Study

10200 1.2 Computer and information sciences

Country of publisher

United States of America

Confidentiality degree

není předmětem státního či obchodního tajemství

Publication form

electronic version available online

References:

URL

RIV identification code

RIV/00216224:14330/20:00114616

Organization unit

Faculty of Informatics

ISBN

978-1-57735-824-4

ISSN

Keywords in English

decision making; Markov decision processes; contextual recommendations

Abstract

V originále

Multiple-environment Markov decision processes (MEMDPs) are MDPs equipped with not one, but multiple probabilistic transition functions, which represent the various possible unknown environments. While the previous research on MEMDPs focused on theoretical properties for long-run average payoff, we study them with discounted-sum payoff and focus on their practical advantages and applications. MEMDPs can be viewed as a special case of Partially observable and Mixed observability MDPs: the state of the system is perfectly observable, but not the environment. We show that the specific structure of MEMDPs allows for more efficient algorithmic analysis, in particular for faster belief updates. We demonstrate the applicability of MEMDPs in several domains. In particular, we formalize the sequential decision-making approach to contextual recommendation systems as MEMDPs and substantially improve over the previous MDP approach.

Links

GA19-15134Y, interní kód MU

Name: Verifikace a analýza pravděpodobnostních programů

Investor: Czech Science Foundation

GJ19-15134Y, research and development project

Name: Verifikace a analýza pravděpodobnostních programů

Citovat

CHATTERJEE, Krishnendu, Martin CHMELÍK, Deep KARKHANIS, Petr NOVOTNÝ and Amélie ROYER. Multiple-Environment Markov Decision Processes: Efficient Analysis and Applications. Online. In Proceedings of the International Conference on Automated Planning and Scheduling. Palo Alto: AAAI Press, 2020, p. 48-56. ISBN 978-1-57735-824-4.

@inproceedings{1718276,
   author = {Chatterjee, Krishnendu and Chmelík, Martin and Karkhanis, Deep and Novotný, Petr and Royer, Amélie},
   address = {Palo Alto},
   booktitle = {Proceedings of the International Conference on Automated Planning and Scheduling},
   keywords = {decision making; Markov decision processes; contextual recommendations},
   howpublished = {elektronická verze "online"},
   language = {eng},
   location = {Palo Alto},
   isbn = {978-1-57735-824-4},
   pages = {48-56},
   publisher = {AAAI Press},
   title = {Multiple-Environment Markov Decision Processes: Efficient Analysis and Applications},
   url = {https://ojs.aaai.org//index.php/ICAPS/article/view/6644},
   year = {2020}
}

TY  - JOUR
ID  - 1718276
AU  - Chatterjee, Krishnendu - Chmelík, Martin - Karkhanis, Deep - Novotný, Petr - Royer, Amélie
PY  - 2020
TI  - Multiple-Environment Markov Decision Processes: Efficient Analysis and Applications
PB  - AAAI Press
CY  - Palo Alto
SN  - 9781577358244
KW  - decision making
KW  - Markov decision processes
KW  - contextual recommendations
UR  - https://ojs.aaai.org//index.php/ICAPS/article/view/6644
L2  - https://ojs.aaai.org//index.php/ICAPS/article/view/6644
N2  - Multiple-environment Markov decision processes (MEMDPs) are MDPs equipped with not one, but multiple probabilistic transition functions, which represent the various possible unknown environments. While the previous research on MEMDPs focused on theoretical properties for long-run average payoff, we study them with discounted-sum payoff and focus on their practical advantages and applications. MEMDPs can be viewed as a special case of Partially observable and Mixed observability MDPs: the state of the system is perfectly observable, but not the environment. We show that the specific structure of MEMDPs allows for more efficient algorithmic analysis, in particular for faster belief updates. We demonstrate the applicability of MEMDPs in several domains. In particular, we formalize the sequential decision-making approach to contextual recommendation systems as MEMDPs and substantially improve over the previous MDP approach.
ER  -

CHATTERJEE, Krishnendu, Martin CHMELÍK, Deep KARKHANIS, Petr NOVOTNÝ and Amélie ROYER. Multiple-Environment Markov Decision Processes: Efficient Analysis and Applications. Online. In \textit{Proceedings of the International Conference on Automated Planning and Scheduling}. Palo Alto: AAAI Press, 2020, p.~48-56. ISBN~978-1-57735-824-4.

Detailed Information on Publication Record