Learning Bayesian networks with thousands of variables

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

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

19 Citaties (Scopus)

Uittreksel

We present a method for learning Bayesian networks from data sets containing thousands of variables without the need for structure constraints. Our approach is made of two parts. The first is a novel algorithm that effectively explores the space of possible parent sets of a node. It guides the exploration towards the most promising parent sets on the basis of an approximated score function that is computed in constant time. The second part is an improvement of an existing ordering-based algorithm for structure optimization. The new algorithm provably achieves a higher score compared to its original formulation. Our novel approach consistently outperforms the state of the art on very large data sets.

Originele taal-2Engels
TitelNIPS'15 Proceedings of the 28th International Conference on Neural Information Processing Systems - Volume 2
Plaats van productieCambridge
UitgeverijMIT Press
Pagina's1864-1872
Aantal pagina's9
StatusGepubliceerd - 2015
Extern gepubliceerdJa
Evenement29th Annual Conference on Neural Information Processing Systems, NIPS 2015 - Montreal, Canada
Duur: 7 dec 201512 dec 2015

Publicatie series

NaamAdvances in Neural Information Processing Systems
ISSN van geprinte versie1049-5258

Congres

Congres29th Annual Conference on Neural Information Processing Systems, NIPS 2015
LandCanada
StadMontreal
Periode7/12/1512/12/15

Vingerafdruk

Bayesian networks

Bibliografische nota

Double blind peer reviewed by multiple reviewers. Acc. rate 22%.

Citeer dit

Scanagatta, M., de Campos, C. P., Corani, G., & Zaffalon, M. (2015). Learning Bayesian networks with thousands of variables. In NIPS'15 Proceedings of the 28th International Conference on Neural Information Processing Systems - Volume 2 (blz. 1864-1872). (Advances in Neural Information Processing Systems). Cambridge: MIT Press.
Scanagatta, Mauro ; de Campos, Cassio P. ; Corani, Giorgio ; Zaffalon, Marco. / Learning Bayesian networks with thousands of variables. NIPS'15 Proceedings of the 28th International Conference on Neural Information Processing Systems - Volume 2. Cambridge : MIT Press, 2015. blz. 1864-1872 (Advances in Neural Information Processing Systems).
@inproceedings{4647fd9406ad48a38adb70de9ec83ab9,
title = "Learning Bayesian networks with thousands of variables",
abstract = "We present a method for learning Bayesian networks from data sets containing thousands of variables without the need for structure constraints. Our approach is made of two parts. The first is a novel algorithm that effectively explores the space of possible parent sets of a node. It guides the exploration towards the most promising parent sets on the basis of an approximated score function that is computed in constant time. The second part is an improvement of an existing ordering-based algorithm for structure optimization. The new algorithm provably achieves a higher score compared to its original formulation. Our novel approach consistently outperforms the state of the art on very large data sets.",
author = "Mauro Scanagatta and {de Campos}, {Cassio P.} and Giorgio Corani and Marco Zaffalon",
note = "Double blind peer reviewed by multiple reviewers. Acc. rate 22{\%}.",
year = "2015",
language = "English",
series = "Advances in Neural Information Processing Systems",
publisher = "MIT Press",
pages = "1864--1872",
booktitle = "NIPS'15 Proceedings of the 28th International Conference on Neural Information Processing Systems - Volume 2",

}

Scanagatta, M, de Campos, CP, Corani, G & Zaffalon, M 2015, Learning Bayesian networks with thousands of variables. in NIPS'15 Proceedings of the 28th International Conference on Neural Information Processing Systems - Volume 2. Advances in Neural Information Processing Systems, MIT Press, Cambridge, blz. 1864-1872, Montreal, Canada, 7/12/15.

Learning Bayesian networks with thousands of variables. / Scanagatta, Mauro; de Campos, Cassio P.; Corani, Giorgio; Zaffalon, Marco.

NIPS'15 Proceedings of the 28th International Conference on Neural Information Processing Systems - Volume 2. Cambridge : MIT Press, 2015. blz. 1864-1872 (Advances in Neural Information Processing Systems).

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

TY - GEN

T1 - Learning Bayesian networks with thousands of variables

AU - Scanagatta, Mauro

AU - de Campos, Cassio P.

AU - Corani, Giorgio

AU - Zaffalon, Marco

N1 - Double blind peer reviewed by multiple reviewers. Acc. rate 22%.

PY - 2015

Y1 - 2015

N2 - We present a method for learning Bayesian networks from data sets containing thousands of variables without the need for structure constraints. Our approach is made of two parts. The first is a novel algorithm that effectively explores the space of possible parent sets of a node. It guides the exploration towards the most promising parent sets on the basis of an approximated score function that is computed in constant time. The second part is an improvement of an existing ordering-based algorithm for structure optimization. The new algorithm provably achieves a higher score compared to its original formulation. Our novel approach consistently outperforms the state of the art on very large data sets.

AB - We present a method for learning Bayesian networks from data sets containing thousands of variables without the need for structure constraints. Our approach is made of two parts. The first is a novel algorithm that effectively explores the space of possible parent sets of a node. It guides the exploration towards the most promising parent sets on the basis of an approximated score function that is computed in constant time. The second part is an improvement of an existing ordering-based algorithm for structure optimization. The new algorithm provably achieves a higher score compared to its original formulation. Our novel approach consistently outperforms the state of the art on very large data sets.

UR - http://www.scopus.com/inward/record.url?scp=84965116806&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84965116806

T3 - Advances in Neural Information Processing Systems

SP - 1864

EP - 1872

BT - NIPS'15 Proceedings of the 28th International Conference on Neural Information Processing Systems - Volume 2

PB - MIT Press

CY - Cambridge

ER -

Scanagatta M, de Campos CP, Corani G, Zaffalon M. Learning Bayesian networks with thousands of variables. In NIPS'15 Proceedings of the 28th International Conference on Neural Information Processing Systems - Volume 2. Cambridge: MIT Press. 2015. blz. 1864-1872. (Advances in Neural Information Processing Systems).