Compiling Relational Bayesian Networks for Exact Inference

Manfred Jaeger; Mark Chavira; Adnan Darwiche

Compiling Relational Bayesian Networks for Exact Inference

Manfred Jaeger, Mark Chavira, Adnan Darwiche

Department of Computer Science

Research output: Contribution to book/anthology/report/conference proceeding › Article in proceeding › Research

381 Downloads (Pure)

Abstract

We describe a system for exact inference with relational Bayesian networks as defined in the publicly available \primula\ tool. The system is based on compiling propositional instances of relational Bayesian networks into arithmetic circuits and then performing online inference by evaluating and differentiating these circuits in time linear in their size. We report on experimental results showing the successful compilation, and efficient inference, on relational Bayesian networks whose {\primula}--generated propositional instances have thousands of variables, and whose jointrees have clusters with hundreds of variables.

Original language	English
Title of host publication	Proceedings of the Second European Workshop on Probabilistic Graphical Models
Editors	P. Lucas (Editor)
Publication date	2004
Publication status	Published - 2004
Event	SECOND EUROPEAN WORKSHOP ON PROBABILISTIC GRAPHICALMODELS 2004 (PGM'04) - Leiden, Netherlands Duration: 4 Oct 2004 → 8 Oct 2004 Conference number: 2

Conference

Conference	SECOND EUROPEAN WORKSHOP ON PROBABILISTIC GRAPHICALMODELS 2004 (PGM'04)
Number	2
Country/Territory	Netherlands
City	Leiden
Period	04/10/2004 → 08/10/2004

Access to Document

Pgm04-finalFinal published version, 114 KB

AUB Link

Search for the material in Aalborg University Library's search engine

Cite this

@inproceedings{22b027409c2e11db8ed6000ea68e967b,

title = "Compiling Relational Bayesian Networks for Exact Inference",

abstract = "We describe a system for exact inference with relational Bayesian networks as defined in the publicly available \primula\ tool. The system is based on compiling propositional instances of relational Bayesian networks into arithmetic circuits and then performing online inference by evaluating and differentiating these circuits in time linear in their size. We report on experimental results showing the successful compilation, and efficient inference, on relational Bayesian networks whose {\primula}--generated propositional instances have thousands of variables, and whose jointrees have clusters with hundreds of variables.",

author = "Manfred Jaeger and Mark Chavira and Adnan Darwiche",

year = "2004",

language = "English",

editor = "(Editor), {P. Lucas}",

booktitle = "Proceedings of the Second European Workshop on Probabilistic Graphical Models",

note = "SECOND EUROPEAN WORKSHOP ON PROBABILISTIC GRAPHICALMODELS 2004 (PGM'04) ; Conference date: 04-10-2004 Through 08-10-2004",

}

TY - GEN

T1 - Compiling Relational Bayesian Networks for Exact Inference

AU - Jaeger, Manfred

AU - Chavira, Mark

AU - Darwiche, Adnan

N1 - Conference code: 2

PY - 2004

Y1 - 2004

N2 - We describe a system for exact inference with relational Bayesian networks as defined in the publicly available \primula\ tool. The system is based on compiling propositional instances of relational Bayesian networks into arithmetic circuits and then performing online inference by evaluating and differentiating these circuits in time linear in their size. We report on experimental results showing the successful compilation, and efficient inference, on relational Bayesian networks whose {\primula}--generated propositional instances have thousands of variables, and whose jointrees have clusters with hundreds of variables.

AB - We describe a system for exact inference with relational Bayesian networks as defined in the publicly available \primula\ tool. The system is based on compiling propositional instances of relational Bayesian networks into arithmetic circuits and then performing online inference by evaluating and differentiating these circuits in time linear in their size. We report on experimental results showing the successful compilation, and efficient inference, on relational Bayesian networks whose {\primula}--generated propositional instances have thousands of variables, and whose jointrees have clusters with hundreds of variables.

M3 - Article in proceeding

BT - Proceedings of the Second European Workshop on Probabilistic Graphical Models

A2 - (Editor), P. Lucas

T2 - SECOND EUROPEAN WORKSHOP ON PROBABILISTIC GRAPHICALMODELS 2004 (PGM'04)

Y2 - 4 October 2004 through 8 October 2004

ER -

Compiling Relational Bayesian Networks for Exact Inference

Abstract

Conference

Access to Document

AUB Link

Fingerprint

Cite this