Exploiting sparsity and sharing in probabilistic sensor data models

S. Evers

Research output: Book/ReportReportProfessional

22 Downloads (Pure)

Abstract

Probabilistic sensor models defined as dynamic Bayesian networks can possess an inherent sparsity that is not reflected in the structure of the network. Classical inference algorithms like variable elimination and junction tree propagation cannot exploit this sparsity. Also, they do not exploit the opportunities for sharing calculations among different time slices of the model. We show that, using a relational representation, inference expressions for these sensor models can be rewritten to make efficient use of sparsity and sharing.
Original languageUndefined
Place of PublicationEnschede
PublisherDatabases (DB)
Number of pages14
Publication statusPublished - 30 Dec 2008

Publication series

NameCTIT Technical Report Series
PublisherCentre for Telematics and Information Technology, University of Twente
No.2008/16200/TR-CTIT-08-68
ISSN (Print)1381-3625

Keywords

  • METIS-255047
  • EWI-14714
  • IR-65252

Cite this

Evers, S. (2008). Exploiting sparsity and sharing in probabilistic sensor data models. (CTIT Technical Report Series; No. 2008/16200/TR-CTIT-08-68). Enschede: Databases (DB).
Evers, S. / Exploiting sparsity and sharing in probabilistic sensor data models. Enschede : Databases (DB), 2008. 14 p. (CTIT Technical Report Series; 2008/16200/TR-CTIT-08-68).
@book{0e6a7e94274448488bd0711964f814fe,
title = "Exploiting sparsity and sharing in probabilistic sensor data models",
abstract = "Probabilistic sensor models defined as dynamic Bayesian networks can possess an inherent sparsity that is not reflected in the structure of the network. Classical inference algorithms like variable elimination and junction tree propagation cannot exploit this sparsity. Also, they do not exploit the opportunities for sharing calculations among different time slices of the model. We show that, using a relational representation, inference expressions for these sensor models can be rewritten to make efficient use of sparsity and sharing.",
keywords = "METIS-255047, EWI-14714, IR-65252",
author = "S. Evers",
year = "2008",
month = "12",
day = "30",
language = "Undefined",
series = "CTIT Technical Report Series",
publisher = "Databases (DB)",
number = "2008/16200/TR-CTIT-08-68",

}

Evers, S 2008, Exploiting sparsity and sharing in probabilistic sensor data models. CTIT Technical Report Series, no. 2008/16200/TR-CTIT-08-68, Databases (DB), Enschede.

Exploiting sparsity and sharing in probabilistic sensor data models. / Evers, S.

Enschede : Databases (DB), 2008. 14 p. (CTIT Technical Report Series; No. 2008/16200/TR-CTIT-08-68).

Research output: Book/ReportReportProfessional

TY - BOOK

T1 - Exploiting sparsity and sharing in probabilistic sensor data models

AU - Evers, S.

PY - 2008/12/30

Y1 - 2008/12/30

N2 - Probabilistic sensor models defined as dynamic Bayesian networks can possess an inherent sparsity that is not reflected in the structure of the network. Classical inference algorithms like variable elimination and junction tree propagation cannot exploit this sparsity. Also, they do not exploit the opportunities for sharing calculations among different time slices of the model. We show that, using a relational representation, inference expressions for these sensor models can be rewritten to make efficient use of sparsity and sharing.

AB - Probabilistic sensor models defined as dynamic Bayesian networks can possess an inherent sparsity that is not reflected in the structure of the network. Classical inference algorithms like variable elimination and junction tree propagation cannot exploit this sparsity. Also, they do not exploit the opportunities for sharing calculations among different time slices of the model. We show that, using a relational representation, inference expressions for these sensor models can be rewritten to make efficient use of sparsity and sharing.

KW - METIS-255047

KW - EWI-14714

KW - IR-65252

M3 - Report

T3 - CTIT Technical Report Series

BT - Exploiting sparsity and sharing in probabilistic sensor data models

PB - Databases (DB)

CY - Enschede

ER -

Evers S. Exploiting sparsity and sharing in probabilistic sensor data models. Enschede: Databases (DB), 2008. 14 p. (CTIT Technical Report Series; 2008/16200/TR-CTIT-08-68).