A hybrid approach to discover Bayesian networks from databases using evolutionary programming

Man Leung WONG, Shing Yan LEE, Kwong Sak LEUNG

Research output: Book Chapters | Papers in Conference ProceedingsConference paper (refereed)Researchpeer-review

17 Citations (Scopus)

Abstract

This paper describes a novel data mining approach that employs evolutionary programming to discover knowledge represented in Bayesian networks. There are two different approaches to the network learning problem. The first one uses dependency analysis, while the second one searches good network structures according to a metric. Unfortunately, both approaches have their own drawbacks. Thus, we propose a novel hybrid algorithm of the two approaches, which consists of two phases, namely, the Conditional Independence (CI) test and the search phases. A new operator is introduced to further enhance the search efficiency. We conduct a number of experiments and compare the hybrid algorithm with our previous algorithm, MDLEP [18], which uses EP for network learning. The empirical results illustrate that the new approach has better performance. We apply the approach to a data sets of direct marketing and compare the performance of the evolved Bayesian networks obtained by the new algorithm with the models generated by other methods. In the comparison, the induced Bayesian networks produced by the new algorithm outperform the other models.
Original languageEnglish
Title of host publicationProceedings - IEEE International Conference on Data Mining, ICDM
Pages498-505
Number of pages8
DOIs
Publication statusPublished - 1 Jan 2002

Fingerprint

Bayesian networks
Evolutionary algorithms
Data mining
Marketing
Experiments

Bibliographical note

Paper presented at the 2nd IEEE International Conference on Data Mining, Dec 09-12, 2002, Maebashi City, Japan. ISBN of the source publication: 9780769517544

Cite this

WONG, M. L., LEE, S. Y., & LEUNG, K. S. (2002). A hybrid approach to discover Bayesian networks from databases using evolutionary programming. In Proceedings - IEEE International Conference on Data Mining, ICDM (pp. 498-505) https://doi.org/10.1109/ICDM.2002.1183994
WONG, Man Leung ; LEE, Shing Yan ; LEUNG, Kwong Sak. / A hybrid approach to discover Bayesian networks from databases using evolutionary programming. Proceedings - IEEE International Conference on Data Mining, ICDM. 2002. pp. 498-505
@inproceedings{e06aa2e61f5a490ca5ee65c1ac9704ba,
title = "A hybrid approach to discover Bayesian networks from databases using evolutionary programming",
abstract = "This paper describes a novel data mining approach that employs evolutionary programming to discover knowledge represented in Bayesian networks. There are two different approaches to the network learning problem. The first one uses dependency analysis, while the second one searches good network structures according to a metric. Unfortunately, both approaches have their own drawbacks. Thus, we propose a novel hybrid algorithm of the two approaches, which consists of two phases, namely, the Conditional Independence (CI) test and the search phases. A new operator is introduced to further enhance the search efficiency. We conduct a number of experiments and compare the hybrid algorithm with our previous algorithm, MDLEP [18], which uses EP for network learning. The empirical results illustrate that the new approach has better performance. We apply the approach to a data sets of direct marketing and compare the performance of the evolved Bayesian networks obtained by the new algorithm with the models generated by other methods. In the comparison, the induced Bayesian networks produced by the new algorithm outperform the other models.",
author = "WONG, {Man Leung} and LEE, {Shing Yan} and LEUNG, {Kwong Sak}",
note = "Paper presented at the 2nd IEEE International Conference on Data Mining, Dec 09-12, 2002, Maebashi City, Japan. ISBN of the source publication: 9780769517544",
year = "2002",
month = "1",
day = "1",
doi = "10.1109/ICDM.2002.1183994",
language = "English",
isbn = "9780769517544",
pages = "498--505",
booktitle = "Proceedings - IEEE International Conference on Data Mining, ICDM",

}

WONG, ML, LEE, SY & LEUNG, KS 2002, A hybrid approach to discover Bayesian networks from databases using evolutionary programming. in Proceedings - IEEE International Conference on Data Mining, ICDM. pp. 498-505. https://doi.org/10.1109/ICDM.2002.1183994

A hybrid approach to discover Bayesian networks from databases using evolutionary programming. / WONG, Man Leung; LEE, Shing Yan; LEUNG, Kwong Sak.

Proceedings - IEEE International Conference on Data Mining, ICDM. 2002. p. 498-505.

Research output: Book Chapters | Papers in Conference ProceedingsConference paper (refereed)Researchpeer-review

TY - GEN

T1 - A hybrid approach to discover Bayesian networks from databases using evolutionary programming

AU - WONG, Man Leung

AU - LEE, Shing Yan

AU - LEUNG, Kwong Sak

N1 - Paper presented at the 2nd IEEE International Conference on Data Mining, Dec 09-12, 2002, Maebashi City, Japan. ISBN of the source publication: 9780769517544

PY - 2002/1/1

Y1 - 2002/1/1

N2 - This paper describes a novel data mining approach that employs evolutionary programming to discover knowledge represented in Bayesian networks. There are two different approaches to the network learning problem. The first one uses dependency analysis, while the second one searches good network structures according to a metric. Unfortunately, both approaches have their own drawbacks. Thus, we propose a novel hybrid algorithm of the two approaches, which consists of two phases, namely, the Conditional Independence (CI) test and the search phases. A new operator is introduced to further enhance the search efficiency. We conduct a number of experiments and compare the hybrid algorithm with our previous algorithm, MDLEP [18], which uses EP for network learning. The empirical results illustrate that the new approach has better performance. We apply the approach to a data sets of direct marketing and compare the performance of the evolved Bayesian networks obtained by the new algorithm with the models generated by other methods. In the comparison, the induced Bayesian networks produced by the new algorithm outperform the other models.

AB - This paper describes a novel data mining approach that employs evolutionary programming to discover knowledge represented in Bayesian networks. There are two different approaches to the network learning problem. The first one uses dependency analysis, while the second one searches good network structures according to a metric. Unfortunately, both approaches have their own drawbacks. Thus, we propose a novel hybrid algorithm of the two approaches, which consists of two phases, namely, the Conditional Independence (CI) test and the search phases. A new operator is introduced to further enhance the search efficiency. We conduct a number of experiments and compare the hybrid algorithm with our previous algorithm, MDLEP [18], which uses EP for network learning. The empirical results illustrate that the new approach has better performance. We apply the approach to a data sets of direct marketing and compare the performance of the evolved Bayesian networks obtained by the new algorithm with the models generated by other methods. In the comparison, the induced Bayesian networks produced by the new algorithm outperform the other models.

UR - http://commons.ln.edu.hk/sw_master/6841

U2 - 10.1109/ICDM.2002.1183994

DO - 10.1109/ICDM.2002.1183994

M3 - Conference paper (refereed)

SN - 9780769517544

SP - 498

EP - 505

BT - Proceedings - IEEE International Conference on Data Mining, ICDM

ER -

WONG ML, LEE SY, LEUNG KS. A hybrid approach to discover Bayesian networks from databases using evolutionary programming. In Proceedings - IEEE International Conference on Data Mining, ICDM. 2002. p. 498-505 https://doi.org/10.1109/ICDM.2002.1183994