PARTIALLY OBSERVABLE MARKOV DECISION PROCESSES WITH PARTIALLY OBSERVABLE RANDOM DISCOUNT FACTORS

E. Everardo Martinez-Garcia; J. Adolfo Minjárez-Sosa; Oscar Vega-Amaya

doi:10.14736/kyb-2022-6-0960

PARTIALLY OBSERVABLE MARKOV DECISION PROCESSES WITH PARTIALLY OBSERVABLE RANDOM DISCOUNT FACTORS

E. Everardo Martinez-Garcia
, J. Adolfo Minjárez-Sosa
, Oscar Vega-Amaya

Research output: Contribution to journal › Article › peer-review

Abstract

This paper deals with a class of partially observable discounted Markov decision processes defined on Borel state and action spaces, under unbounded one-stage cost. The discount rate is a stochastic process evolving according to a difference equation, which is also assumed to be partially observable. Introducing a suitable control model and filtering processes, we prove the existence of optimal control policies. In addition, we illustrate our results in a class of GI/GI/1 queueing systems where we obtain explicitly the corresponding optimality equation and the filtering process.

Original language	English
Pages (from-to)	960-983
Number of pages	24
Journal	Kybernetika
Volume	58
Issue number	6
DOIs	https://doi.org/10.14736/kyb-2022-6-0960
State	Published - 2022

Bibliographical note

Publisher Copyright:
© 2022 Institute of Information Theory and Automation of The Czech Academy of Sciences. All rights reserved.

Keywords

discounted criterion
optimal policies
partially observable systems
queueing models
random discount factors

Access to Document

10.14736/kyb-2022-6-0960

Cite this

@article{ac104b30e51c4c11aadbc1e9b18ec3e0,

title = "PARTIALLY OBSERVABLE MARKOV DECISION PROCESSES WITH PARTIALLY OBSERVABLE RANDOM DISCOUNT FACTORS",

abstract = "This paper deals with a class of partially observable discounted Markov decision processes defined on Borel state and action spaces, under unbounded one-stage cost. The discount rate is a stochastic process evolving according to a difference equation, which is also assumed to be partially observable. Introducing a suitable control model and filtering processes, we prove the existence of optimal control policies. In addition, we illustrate our results in a class of GI/GI/1 queueing systems where we obtain explicitly the corresponding optimality equation and the filtering process.",

keywords = "discounted criterion, optimal policies, partially observable systems, queueing models, random discount factors",

author = "Martinez-Garcia, \{E. Everardo\} and Minj{\'a}rez-Sosa, \{J. Adolfo\} and Oscar Vega-Amaya",

year = "2022",

doi = "10.14736/kyb-2022-6-0960",

language = "Ingl{\'e}s",

volume = "58",

pages = "960--983",

journal = "Kybernetika",

issn = "0023-5954",

publisher = "Institute of Information Theory and Automation of The Czech Academy of Sciences",

number = "6",

}

TY - JOUR

T1 - PARTIALLY OBSERVABLE MARKOV DECISION PROCESSES WITH PARTIALLY OBSERVABLE RANDOM DISCOUNT FACTORS

AU - Martinez-Garcia, E. Everardo

AU - Minjárez-Sosa, J. Adolfo

AU - Vega-Amaya, Oscar

PY - 2022

Y1 - 2022

N2 - This paper deals with a class of partially observable discounted Markov decision processes defined on Borel state and action spaces, under unbounded one-stage cost. The discount rate is a stochastic process evolving according to a difference equation, which is also assumed to be partially observable. Introducing a suitable control model and filtering processes, we prove the existence of optimal control policies. In addition, we illustrate our results in a class of GI/GI/1 queueing systems where we obtain explicitly the corresponding optimality equation and the filtering process.

AB - This paper deals with a class of partially observable discounted Markov decision processes defined on Borel state and action spaces, under unbounded one-stage cost. The discount rate is a stochastic process evolving according to a difference equation, which is also assumed to be partially observable. Introducing a suitable control model and filtering processes, we prove the existence of optimal control policies. In addition, we illustrate our results in a class of GI/GI/1 queueing systems where we obtain explicitly the corresponding optimality equation and the filtering process.

KW - discounted criterion

KW - optimal policies

KW - partially observable systems

KW - queueing models

KW - random discount factors

UR - https://www.scopus.com/pages/publications/85161883948

U2 - 10.14736/kyb-2022-6-0960

DO - 10.14736/kyb-2022-6-0960

M3 - Artículo

AN - SCOPUS:85161883948

SN - 0023-5954

VL - 58

SP - 960

EP - 983

JO - Kybernetika

JF - Kybernetika

IS - 6

ER -

PARTIALLY OBSERVABLE MARKOV DECISION PROCESSES WITH PARTIALLY OBSERVABLE RANDOM DISCOUNT FACTORS

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this