SparsityBoost: A new scoring function for learning Bayesian network structure

Eliot Brenner, David Sontag

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We give a new consistent scoring function for structure learning of Bayesian networks. In contrast to traditional approaches to score-based structure learning, such as BDeu or MDL, the complexity penalty that we propose is data-dependent and is given by the probability that a conditional independence test correctly shows that an edge cannot exist. What really distinguishes this new scoring function from earlier work is that it has the property of becoming computationally easier to maximize as the amount of data increases. We prove a polynomial sample complexity result, showing that maximizing this score is guaranteed to correctly learn a structure with no false edges and a distribution close to the generating distribution, whenever there exists a Bayesian network which is a perfect map for the data generating distribution. Although the new score can be used with any search algorithm, we give empirical results showing that it is particularly effective when used together with a linear programming relaxation approach to Bayesian network structure learning.

Original languageEnglish (US)
Title of host publicationUncertainty in Artificial Intelligence - Proceedings of the 29th Conference, UAI 2013
Pages112-121
Number of pages10
StatePublished - 2013
Event29th Conference on Uncertainty in Artificial Intelligence, UAI 2013 - Bellevue, WA, United States
Duration: Jul 11 2013Jul 15 2013

Other

Other29th Conference on Uncertainty in Artificial Intelligence, UAI 2013
CountryUnited States
CityBellevue, WA
Period7/11/137/15/13

Fingerprint

Bayesian networks
Linear programming
Polynomials

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Brenner, E., & Sontag, D. (2013). SparsityBoost: A new scoring function for learning Bayesian network structure. In Uncertainty in Artificial Intelligence - Proceedings of the 29th Conference, UAI 2013 (pp. 112-121)

SparsityBoost : A new scoring function for learning Bayesian network structure. / Brenner, Eliot; Sontag, David.

Uncertainty in Artificial Intelligence - Proceedings of the 29th Conference, UAI 2013. 2013. p. 112-121.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Brenner, E & Sontag, D 2013, SparsityBoost: A new scoring function for learning Bayesian network structure. in Uncertainty in Artificial Intelligence - Proceedings of the 29th Conference, UAI 2013. pp. 112-121, 29th Conference on Uncertainty in Artificial Intelligence, UAI 2013, Bellevue, WA, United States, 7/11/13.
Brenner E, Sontag D. SparsityBoost: A new scoring function for learning Bayesian network structure. In Uncertainty in Artificial Intelligence - Proceedings of the 29th Conference, UAI 2013. 2013. p. 112-121
Brenner, Eliot ; Sontag, David. / SparsityBoost : A new scoring function for learning Bayesian network structure. Uncertainty in Artificial Intelligence - Proceedings of the 29th Conference, UAI 2013. 2013. pp. 112-121
@inproceedings{256253f07cbf4904999af897164c4341,
title = "SparsityBoost: A new scoring function for learning Bayesian network structure",
abstract = "We give a new consistent scoring function for structure learning of Bayesian networks. In contrast to traditional approaches to score-based structure learning, such as BDeu or MDL, the complexity penalty that we propose is data-dependent and is given by the probability that a conditional independence test correctly shows that an edge cannot exist. What really distinguishes this new scoring function from earlier work is that it has the property of becoming computationally easier to maximize as the amount of data increases. We prove a polynomial sample complexity result, showing that maximizing this score is guaranteed to correctly learn a structure with no false edges and a distribution close to the generating distribution, whenever there exists a Bayesian network which is a perfect map for the data generating distribution. Although the new score can be used with any search algorithm, we give empirical results showing that it is particularly effective when used together with a linear programming relaxation approach to Bayesian network structure learning.",
author = "Eliot Brenner and David Sontag",
year = "2013",
language = "English (US)",
pages = "112--121",
booktitle = "Uncertainty in Artificial Intelligence - Proceedings of the 29th Conference, UAI 2013",

}

TY - GEN

T1 - SparsityBoost

T2 - A new scoring function for learning Bayesian network structure

AU - Brenner, Eliot

AU - Sontag, David

PY - 2013

Y1 - 2013

N2 - We give a new consistent scoring function for structure learning of Bayesian networks. In contrast to traditional approaches to score-based structure learning, such as BDeu or MDL, the complexity penalty that we propose is data-dependent and is given by the probability that a conditional independence test correctly shows that an edge cannot exist. What really distinguishes this new scoring function from earlier work is that it has the property of becoming computationally easier to maximize as the amount of data increases. We prove a polynomial sample complexity result, showing that maximizing this score is guaranteed to correctly learn a structure with no false edges and a distribution close to the generating distribution, whenever there exists a Bayesian network which is a perfect map for the data generating distribution. Although the new score can be used with any search algorithm, we give empirical results showing that it is particularly effective when used together with a linear programming relaxation approach to Bayesian network structure learning.

AB - We give a new consistent scoring function for structure learning of Bayesian networks. In contrast to traditional approaches to score-based structure learning, such as BDeu or MDL, the complexity penalty that we propose is data-dependent and is given by the probability that a conditional independence test correctly shows that an edge cannot exist. What really distinguishes this new scoring function from earlier work is that it has the property of becoming computationally easier to maximize as the amount of data increases. We prove a polynomial sample complexity result, showing that maximizing this score is guaranteed to correctly learn a structure with no false edges and a distribution close to the generating distribution, whenever there exists a Bayesian network which is a perfect map for the data generating distribution. Although the new score can be used with any search algorithm, we give empirical results showing that it is particularly effective when used together with a linear programming relaxation approach to Bayesian network structure learning.

UR - http://www.scopus.com/inward/record.url?scp=84888177974&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84888177974&partnerID=8YFLogxK

M3 - Conference contribution

SP - 112

EP - 121

BT - Uncertainty in Artificial Intelligence - Proceedings of the 29th Conference, UAI 2013

ER -