Learning treewidth-bounded Bayesian networks with thousands of variables

Mauro Scanagatta; Giorgio Corani; Cassio P. de Campos; Marco Zaffalon

Learning treewidth-bounded Bayesian networks with thousands of variables

Mauro Scanagatta, Giorgio Corani, Cassio P. de Campos, Marco Zaffalon

School of Electronics, Electrical Engineering and Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Other chapter contribution

34 Citations (Scopus)

874 Downloads (Pure)

Abstract

We present a method for learning treewidth-bounded Bayesian networks from data sets containing thousands of variables. Bounding the treewidth of a Bayesian network greatly reduces the complexity of inferences. Yet, being a global property of the graph, it considerably increases the difficulty of the learning process. Our novel algorithm accomplishes this task, scaling both to large domains and to large treewidths. Our novel approach consistently outperforms the state of the art on experiments with up to thousands of variables.

Original language	English
Title of host publication	Proceedings of the 30th Conference on Neutral Information Processing Systems, NIPS 2016
Editors	D. Lee, M. Sugiyama, U. Luxburg, I. Guyon, R. Garnett
Publisher	NIPS Foundation, Inc
Number of pages	9
ISBN (Electronic)	9781510838819
Publication status	Published - 12 Aug 2016
Event	Annual Conference on Neural Information Processing Systems 2016 - Centre Convencions Internacional Barcelona, Barcelona, Spain Duration: 05 Dec 2016 → 10 Dec 2016 http://nips.cc

Publication series

Name	Advances in Neural Information Processing Systems
Volume	29

Conference

Conference	Annual Conference on Neural Information Processing Systems 2016
Abbreviated title	NIPS 2016
Country/Territory	Spain
City	Barcelona
Period	05/12/2016 → 10/12/2016
Internet address	http://nips.cc

Access to Document

Learning treewidth-bounded Bayesian networks with thousands of variables
Copyright 2016 The Authors. This work is made available online in accordance with the publisher’s policies. Please refer to any applicable terms of use of the publisher.
Accepted author manuscript, 258 KBLicence: Unspecified

Cite this

Scanagatta, M., Corani, G., de Campos, C. P., & Zaffalon, M. (2016). Learning treewidth-bounded Bayesian networks with thousands of variables. In D. Lee, M. Sugiyama, U. Luxburg, I. Guyon, & R. Garnett (Eds.), Proceedings of the 30th Conference on Neutral Information Processing Systems, NIPS 2016 (Advances in Neural Information Processing Systems; Vol. 29). NIPS Foundation, Inc.

Scanagatta, Mauro ; Corani, Giorgio ; de Campos, Cassio P. et al. / Learning treewidth-bounded Bayesian networks with thousands of variables. Proceedings of the 30th Conference on Neutral Information Processing Systems, NIPS 2016. editor / D. Lee ; M. Sugiyama ; U. Luxburg ; I. Guyon ; R. Garnett. NIPS Foundation, Inc, 2016. (Advances in Neural Information Processing Systems).

@inbook{e836c83fb5cc469cb3040ddd6be4bd0f,

title = "Learning treewidth-bounded Bayesian networks with thousands of variables",

abstract = "We present a method for learning treewidth-bounded Bayesian networks from data sets containing thousands of variables. Bounding the treewidth of a Bayesian network greatly reduces the complexity of inferences. Yet, being a global property of the graph, it considerably increases the difficulty of the learning process. Our novel algorithm accomplishes this task, scaling both to large domains and to large treewidths. Our novel approach consistently outperforms the state of the art on experiments with up to thousands of variables.",

author = "Mauro Scanagatta and Giorgio Corani and {de Campos}, {Cassio P.} and Marco Zaffalon",

year = "2016",

month = aug,

day = "12",

language = "English",

series = "Advances in Neural Information Processing Systems",

publisher = "NIPS Foundation, Inc",

editor = "D. Lee and M. Sugiyama and U. Luxburg and I. Guyon and R. Garnett",

booktitle = "Proceedings of the 30th Conference on Neutral Information Processing Systems, NIPS 2016",

note = "Annual Conference on Neural Information Processing Systems 2016, NIPS 2016 ; Conference date: 05-12-2016 Through 10-12-2016",

url = "http://nips.cc",

}

Scanagatta, M, Corani, G, de Campos, CP & Zaffalon, M 2016, Learning treewidth-bounded Bayesian networks with thousands of variables. in D Lee, M Sugiyama, U Luxburg, I Guyon & R Garnett (eds), Proceedings of the 30th Conference on Neutral Information Processing Systems, NIPS 2016. Advances in Neural Information Processing Systems, vol. 29, NIPS Foundation, Inc, Annual Conference on Neural Information Processing Systems 2016, Barcelona, Spain, 05/12/2016.

Learning treewidth-bounded Bayesian networks with thousands of variables. / Scanagatta, Mauro; Corani, Giorgio; de Campos, Cassio P. et al.
Proceedings of the 30th Conference on Neutral Information Processing Systems, NIPS 2016. ed. / D. Lee; M. Sugiyama; U. Luxburg; I. Guyon; R. Garnett. NIPS Foundation, Inc, 2016. (Advances in Neural Information Processing Systems; Vol. 29).

Research output: Chapter in Book/Report/Conference proceeding › Other chapter contribution

TY - CHAP

T1 - Learning treewidth-bounded Bayesian networks with thousands of variables

AU - Scanagatta, Mauro

AU - Corani, Giorgio

AU - de Campos, Cassio P.

AU - Zaffalon, Marco

PY - 2016/8/12

Y1 - 2016/8/12

N2 - We present a method for learning treewidth-bounded Bayesian networks from data sets containing thousands of variables. Bounding the treewidth of a Bayesian network greatly reduces the complexity of inferences. Yet, being a global property of the graph, it considerably increases the difficulty of the learning process. Our novel algorithm accomplishes this task, scaling both to large domains and to large treewidths. Our novel approach consistently outperforms the state of the art on experiments with up to thousands of variables.

AB - We present a method for learning treewidth-bounded Bayesian networks from data sets containing thousands of variables. Bounding the treewidth of a Bayesian network greatly reduces the complexity of inferences. Yet, being a global property of the graph, it considerably increases the difficulty of the learning process. Our novel algorithm accomplishes this task, scaling both to large domains and to large treewidths. Our novel approach consistently outperforms the state of the art on experiments with up to thousands of variables.

M3 - Other chapter contribution

T3 - Advances in Neural Information Processing Systems

BT - Proceedings of the 30th Conference on Neutral Information Processing Systems, NIPS 2016

A2 - Lee, D.

A2 - Sugiyama, M.

A2 - Luxburg, U.

A2 - Guyon, I.

A2 - Garnett, R.

PB - NIPS Foundation, Inc

T2 - Annual Conference on Neural Information Processing Systems 2016

Y2 - 5 December 2016 through 10 December 2016

ER -

Learning treewidth-bounded Bayesian networks with thousands of variables

Abstract

Publication series

Conference

Access to Document

Fingerprint

Cite this