The policy gradient estimation for continuous-time partially observable Markovian decision processes

Bo Tang; Yan Jie Li; Bao Qun Yin

The policy gradient estimation for continuous-time partially observable Markovian decision processes

Bo Tang^*
, Yan Jie Li
, Bao Qun Yin

^*Corresponding author for this work

University of Science and Technology of China

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

Abstract

An algorithm for estimating the policy gradient is presented for the performance optimization of continuous-time partially observable Markovian decision processes (CTPOMDPs). This estimation algorithm is obtained by extending the corresponding estimation algorithm for discrete-time partially observable Markovian decision processes (DTPOMDP's), using the conformity method. The convergence and the error bound of this algorithm are analyzed; and a numerical example is provided to illustrate its application.

Original language	English
Pages (from-to)	805-808
Number of pages	4
Journal	Kongzhi Lilun Yu Yingyong/Control Theory and Applications
Volume	26
Issue number	7
State	Published - Jul 2009
Externally published	Yes

Keywords

CTPOMDP
Conformity
Error bound
Policy gradient estimation

Cite this

@article{eae57f70b1054446902ce49c062f0ebd,

title = "The policy gradient estimation for continuous-time partially observable Markovian decision processes",

abstract = "An algorithm for estimating the policy gradient is presented for the performance optimization of continuous-time partially observable Markovian decision processes (CTPOMDPs). This estimation algorithm is obtained by extending the corresponding estimation algorithm for discrete-time partially observable Markovian decision processes (DTPOMDP's), using the conformity method. The convergence and the error bound of this algorithm are analyzed; and a numerical example is provided to illustrate its application.",

keywords = "CTPOMDP, Conformity, Error bound, Policy gradient estimation",

author = "Bo Tang and Li, \{Yan Jie\} and Yin, \{Bao Qun\}",

year = "2009",

month = jul,

language = "英语",

volume = "26",

pages = "805--808",

journal = "Kongzhi Lilun Yu Yingyong/Control Theory and Applications",

issn = "1000-8152",

publisher = "South China University of Technology",

number = "7",

}

TY - JOUR

T1 - The policy gradient estimation for continuous-time partially observable Markovian decision processes

AU - Tang, Bo

AU - Li, Yan Jie

AU - Yin, Bao Qun

PY - 2009/7

Y1 - 2009/7

N2 - An algorithm for estimating the policy gradient is presented for the performance optimization of continuous-time partially observable Markovian decision processes (CTPOMDPs). This estimation algorithm is obtained by extending the corresponding estimation algorithm for discrete-time partially observable Markovian decision processes (DTPOMDP's), using the conformity method. The convergence and the error bound of this algorithm are analyzed; and a numerical example is provided to illustrate its application.

AB - An algorithm for estimating the policy gradient is presented for the performance optimization of continuous-time partially observable Markovian decision processes (CTPOMDPs). This estimation algorithm is obtained by extending the corresponding estimation algorithm for discrete-time partially observable Markovian decision processes (DTPOMDP's), using the conformity method. The convergence and the error bound of this algorithm are analyzed; and a numerical example is provided to illustrate its application.

KW - CTPOMDP

KW - Conformity

KW - Error bound

KW - Policy gradient estimation

UR - https://www.scopus.com/pages/publications/69749091027

M3 - 文章

AN - SCOPUS:69749091027

SN - 1000-8152

VL - 26

SP - 805

EP - 808

JO - Kongzhi Lilun Yu Yingyong/Control Theory and Applications

JF - Kongzhi Lilun Yu Yingyong/Control Theory and Applications

IS - 7

ER -

The policy gradient estimation for continuous-time partially observable Markovian decision processes

Abstract

Keywords

Other files and links

Fingerprint

Cite this