Evaluating point-based POMDP solvers on multicore machines

Guy Shani

doi:10.1109/TSMCB.2009.2034015

Evaluating point-based POMDP solvers on multicore machines

Guy Shani

Department of Software and Information Systems Engineering

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

11 Scopus citations

Abstract

Recent scaling up of partially observable Markov decision process solvers toward realistic applications is largely due to point-based methods which quickly provide approximate solutions for midsized problems. New multicore machines offer an opportunity to scale up to larger domains. These machines support parallel execution and can speed up existing algorithms considerably. In this paper, we evaluate several ways in which point-based algorithms can be adapted to parallel computing. We overview the challenges and opportunities and present experimental results, providing evidence to the usability of our suggestions.

Original language	English
Article number	5332315
Pages (from-to)	1062-1074
Number of pages	13
Journal	IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics
Volume	40
Issue number	4
DOIs	https://doi.org/10.1109/TSMCB.2009.2034015
State	Published - 1 Aug 2010

Keywords

Multi-core machines
parallel computing
partially observable Markov decision processes (POMDP)
point-based value iteration

ASJC Scopus subject areas

Control and Systems Engineering
Software
Information Systems
Human-Computer Interaction
Computer Science Applications
Electrical and Electronic Engineering

Access to Document

10.1109/TSMCB.2009.2034015

Cite this

@article{068f42f374184f23bee503910ba49f69,

title = "Evaluating point-based POMDP solvers on multicore machines",

abstract = "Recent scaling up of partially observable Markov decision process solvers toward realistic applications is largely due to point-based methods which quickly provide approximate solutions for midsized problems. New multicore machines offer an opportunity to scale up to larger domains. These machines support parallel execution and can speed up existing algorithms considerably. In this paper, we evaluate several ways in which point-based algorithms can be adapted to parallel computing. We overview the challenges and opportunities and present experimental results, providing evidence to the usability of our suggestions.",

keywords = "Multi-core machines, parallel computing, partially observable Markov decision processes (POMDP), point-based value iteration",

author = "Guy Shani",

year = "2010",

month = aug,

day = "1",

doi = "10.1109/TSMCB.2009.2034015",

language = "English",

volume = "40",

pages = "1062--1074",

journal = "IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics",

issn = "1083-4419",

publisher = "Institute of Electrical and Electronics Engineers",

number = "4",

}

TY - JOUR

T1 - Evaluating point-based POMDP solvers on multicore machines

AU - Shani, Guy

PY - 2010/8/1

Y1 - 2010/8/1

N2 - Recent scaling up of partially observable Markov decision process solvers toward realistic applications is largely due to point-based methods which quickly provide approximate solutions for midsized problems. New multicore machines offer an opportunity to scale up to larger domains. These machines support parallel execution and can speed up existing algorithms considerably. In this paper, we evaluate several ways in which point-based algorithms can be adapted to parallel computing. We overview the challenges and opportunities and present experimental results, providing evidence to the usability of our suggestions.

AB - Recent scaling up of partially observable Markov decision process solvers toward realistic applications is largely due to point-based methods which quickly provide approximate solutions for midsized problems. New multicore machines offer an opportunity to scale up to larger domains. These machines support parallel execution and can speed up existing algorithms considerably. In this paper, we evaluate several ways in which point-based algorithms can be adapted to parallel computing. We overview the challenges and opportunities and present experimental results, providing evidence to the usability of our suggestions.

KW - Multi-core machines

KW - parallel computing

KW - partially observable Markov decision processes (POMDP)

KW - point-based value iteration

UR - http://www.scopus.com/inward/record.url?scp=77954758939&partnerID=8YFLogxK

U2 - 10.1109/TSMCB.2009.2034015

DO - 10.1109/TSMCB.2009.2034015

M3 - Article

AN - SCOPUS:77954758939

SN - 1083-4419

VL - 40

SP - 1062

EP - 1074

JO - IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics

JF - IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics

IS - 4

M1 - 5332315

ER -

Evaluating point-based POMDP solvers on multicore machines

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this