A Review of Inference Algorithms for Hybrid Bayesian Networks

Antonio Salmerón, Rafael Rumí, Helge Langseth, Thomas Dyhre Nielsen, Anders Læsø Madsen

Research output: Contribution to journalReview articleResearchpeer-review

1 Citation (Scopus)
88 Downloads (Pure)

Abstract

Hybrid Bayesian networks have received an increasing attention during the last years. The difference with respect to standard Bayesian networks is that they can host discrete and continuous variables simultaneously, which extends the applicability of the Bayesian network framework in general. However, this extra feature also comes at a cost: inference in these types of models is computationally more challenging and the underlying models and updating procedures may not even support closed-form solutions. In this paper we provide an overview of the main trends and principled approaches for performing inference in hybrid Bayesian networks. The methods covered in the paper are organized and discussed according to their methodological basis. We consider how the methods have been extended and adapted to also include (hybrid) dynamic Bayesian networks, and we end with an overview of established software systems supporting inference in these types of models.
Original languageEnglish
JournalJournal of Artificial Intelligence Research
Volume62
Pages (from-to)799-828
Number of pages30
ISSN1076-9757
DOIs
Publication statusPublished - 1 Aug 2018

Fingerprint

Bayesian networks
Costs

Cite this

@article{9727b04668c2441eac324ab9b2184ff8,
title = "A Review of Inference Algorithms for Hybrid Bayesian Networks",
abstract = "Hybrid Bayesian networks have received an increasing attention during the last years. The difference with respect to standard Bayesian networks is that they can host discrete and continuous variables simultaneously, which extends the applicability of the Bayesian network framework in general. However, this extra feature also comes at a cost: inference in these types of models is computationally more challenging and the underlying models and updating procedures may not even support closed-form solutions. In this paper we provide an overview of the main trends and principled approaches for performing inference in hybrid Bayesian networks. The methods covered in the paper are organized and discussed according to their methodological basis. We consider how the methods have been extended and adapted to also include (hybrid) dynamic Bayesian networks, and we end with an overview of established software systems supporting inference in these types of models.",
author = "Antonio Salmer{\'o}n and Rafael Rum{\'i} and Helge Langseth and Nielsen, {Thomas Dyhre} and Madsen, {Anders L{\ae}s{\o}}",
year = "2018",
month = "8",
day = "1",
doi = "10.1613/jair.1.11228",
language = "English",
volume = "62",
pages = "799--828",
journal = "Journal of Artificial Intelligence Research",
issn = "1076-9757",
publisher = "A A A I Press",

}

A Review of Inference Algorithms for Hybrid Bayesian Networks. / Salmerón, Antonio; Rumí, Rafael; Langseth, Helge; Nielsen, Thomas Dyhre; Madsen, Anders Læsø.

In: Journal of Artificial Intelligence Research, Vol. 62, 01.08.2018, p. 799-828.

Research output: Contribution to journalReview articleResearchpeer-review

TY - JOUR

T1 - A Review of Inference Algorithms for Hybrid Bayesian Networks

AU - Salmerón, Antonio

AU - Rumí, Rafael

AU - Langseth, Helge

AU - Nielsen, Thomas Dyhre

AU - Madsen, Anders Læsø

PY - 2018/8/1

Y1 - 2018/8/1

N2 - Hybrid Bayesian networks have received an increasing attention during the last years. The difference with respect to standard Bayesian networks is that they can host discrete and continuous variables simultaneously, which extends the applicability of the Bayesian network framework in general. However, this extra feature also comes at a cost: inference in these types of models is computationally more challenging and the underlying models and updating procedures may not even support closed-form solutions. In this paper we provide an overview of the main trends and principled approaches for performing inference in hybrid Bayesian networks. The methods covered in the paper are organized and discussed according to their methodological basis. We consider how the methods have been extended and adapted to also include (hybrid) dynamic Bayesian networks, and we end with an overview of established software systems supporting inference in these types of models.

AB - Hybrid Bayesian networks have received an increasing attention during the last years. The difference with respect to standard Bayesian networks is that they can host discrete and continuous variables simultaneously, which extends the applicability of the Bayesian network framework in general. However, this extra feature also comes at a cost: inference in these types of models is computationally more challenging and the underlying models and updating procedures may not even support closed-form solutions. In this paper we provide an overview of the main trends and principled approaches for performing inference in hybrid Bayesian networks. The methods covered in the paper are organized and discussed according to their methodological basis. We consider how the methods have been extended and adapted to also include (hybrid) dynamic Bayesian networks, and we end with an overview of established software systems supporting inference in these types of models.

UR - http://www.scopus.com/inward/record.url?scp=85052902861&partnerID=8YFLogxK

U2 - 10.1613/jair.1.11228

DO - 10.1613/jair.1.11228

M3 - Review article

VL - 62

SP - 799

EP - 828

JO - Journal of Artificial Intelligence Research

JF - Journal of Artificial Intelligence Research

SN - 1076-9757

ER -