Mining Bayesian networks from direct marketing databases with missing values

Yuan Yuan GUO, Man Leung WONG

Research output: Book Chapters | Papers in Conference ProceedingsBook ChapterResearchpeer-review

Abstract

Discovering knowledge from huge databases with missing values is a challenging problem in Data Mining. In this paper, a novel hybrid algorithm for learning knowledge represented in Bayesian Networks is discussed. The new algorithm combines an evolutionary algorithm with the Expectation-Maximization (EM) algorithm to overcome the problem of getting stuck in sub-optimal solutions which occurs in most existing learning algorithms. The experimental results on the databases generated from several benchmark network structures illustrate that our system outperforms some state-of-the-art algorithms. We also apply our system to a direct marketing problem, and compare the performance of the discovered Bayesian networks with the response models obtained by other algorithms. In the comparison, the Bayesian networks learned by our system outperform others.
Original languageEnglish
Title of host publicationIntelligent and evolutionary systems
PublisherSpringer-Verlag
Pages13-35
Number of pages23
ISBN (Print)9783540959779
DOIs
Publication statusPublished - 1 Jan 2009

Fingerprint

Bayesian networks
Marketing
Evolutionary algorithms
Learning algorithms
Data mining

Cite this

GUO, Y. Y., & WONG, M. L. (2009). Mining Bayesian networks from direct marketing databases with missing values. In Intelligent and evolutionary systems (pp. 13-35). Springer-Verlag. https://doi.org/10.1007/978-3-540-95978-6_2
GUO, Yuan Yuan ; WONG, Man Leung. / Mining Bayesian networks from direct marketing databases with missing values. Intelligent and evolutionary systems. Springer-Verlag, 2009. pp. 13-35
@inbook{f43820c871b44efcb93e6c63997200ba,
title = "Mining Bayesian networks from direct marketing databases with missing values",
abstract = "Discovering knowledge from huge databases with missing values is a challenging problem in Data Mining. In this paper, a novel hybrid algorithm for learning knowledge represented in Bayesian Networks is discussed. The new algorithm combines an evolutionary algorithm with the Expectation-Maximization (EM) algorithm to overcome the problem of getting stuck in sub-optimal solutions which occurs in most existing learning algorithms. The experimental results on the databases generated from several benchmark network structures illustrate that our system outperforms some state-of-the-art algorithms. We also apply our system to a direct marketing problem, and compare the performance of the discovered Bayesian networks with the response models obtained by other algorithms. In the comparison, the Bayesian networks learned by our system outperform others.",
author = "GUO, {Yuan Yuan} and WONG, {Man Leung}",
year = "2009",
month = "1",
day = "1",
doi = "10.1007/978-3-540-95978-6_2",
language = "English",
isbn = "9783540959779",
pages = "13--35",
booktitle = "Intelligent and evolutionary systems",
publisher = "Springer-Verlag",

}

Mining Bayesian networks from direct marketing databases with missing values. / GUO, Yuan Yuan; WONG, Man Leung.

Intelligent and evolutionary systems. Springer-Verlag, 2009. p. 13-35.

Research output: Book Chapters | Papers in Conference ProceedingsBook ChapterResearchpeer-review

TY - CHAP

T1 - Mining Bayesian networks from direct marketing databases with missing values

AU - GUO, Yuan Yuan

AU - WONG, Man Leung

PY - 2009/1/1

Y1 - 2009/1/1

N2 - Discovering knowledge from huge databases with missing values is a challenging problem in Data Mining. In this paper, a novel hybrid algorithm for learning knowledge represented in Bayesian Networks is discussed. The new algorithm combines an evolutionary algorithm with the Expectation-Maximization (EM) algorithm to overcome the problem of getting stuck in sub-optimal solutions which occurs in most existing learning algorithms. The experimental results on the databases generated from several benchmark network structures illustrate that our system outperforms some state-of-the-art algorithms. We also apply our system to a direct marketing problem, and compare the performance of the discovered Bayesian networks with the response models obtained by other algorithms. In the comparison, the Bayesian networks learned by our system outperform others.

AB - Discovering knowledge from huge databases with missing values is a challenging problem in Data Mining. In this paper, a novel hybrid algorithm for learning knowledge represented in Bayesian Networks is discussed. The new algorithm combines an evolutionary algorithm with the Expectation-Maximization (EM) algorithm to overcome the problem of getting stuck in sub-optimal solutions which occurs in most existing learning algorithms. The experimental results on the databases generated from several benchmark network structures illustrate that our system outperforms some state-of-the-art algorithms. We also apply our system to a direct marketing problem, and compare the performance of the discovered Bayesian networks with the response models obtained by other algorithms. In the comparison, the Bayesian networks learned by our system outperform others.

UR - http://commons.ln.edu.hk/sw_master/2333

U2 - 10.1007/978-3-540-95978-6_2

DO - 10.1007/978-3-540-95978-6_2

M3 - Book Chapter

SN - 9783540959779

SP - 13

EP - 35

BT - Intelligent and evolutionary systems

PB - Springer-Verlag

ER -