Diversity in combinations of heterogeneous classifiers

Kuo Wei Hsu, Jaideep Srivastava

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

14 Citations (Scopus)

Abstract

In this paper, we introduce the use of combinations of heterogeneous classifiers to achieve better diversity. Conducting theoretical and empirical analyses of the diversity of combinations of heterogeneous classifiers, we study the relationship between heterogeneity and diversity. On the one hand, the theoretical analysis serves as a foundation for employing heterogeneous classifiers in Multi-Classifier Systems or ensembles. On the other hand, experimental results provide empirical evidence. We consider synthetic as well as real data sets, utilize classification algorithms that are essentially different, and employ various popular diversity measures for evaluation. Two interesting observations will contribute to the future design of Multi-Classifier Systems and ensemble techniques. First, the diversity among heterogeneous classifiers is higher than that among homogeneous ones, and hence using heterogeneous classifiers to construct classifier combinations would increase the diversity. Second, the heterogeneity primarily results from different classification algorithms rather than the same algorithm with different parameters.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages923-932
Number of pages10
Volume5476 LNAI
DOIs
Publication statusPublished - 2009
Externally publishedYes
Event13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2009 - Bangkok, Thailand
Duration: 27 Apr 200930 Apr 2009

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5476 LNAI
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2009
CountryThailand
CityBangkok
Period27/4/0930/4/09

Fingerprint

Classifiers
Classifier
Classification Algorithm
Ensemble
Classifier Combination
Theoretical Analysis
Evaluation
Experimental Results

Keywords

  • Diversity
  • Ensemble
  • Heterogeneity
  • Multi-classifier system

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Hsu, K. W., & Srivastava, J. (2009). Diversity in combinations of heterogeneous classifiers. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5476 LNAI, pp. 923-932). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5476 LNAI). https://doi.org/10.1007/978-3-642-01307-2_97

Diversity in combinations of heterogeneous classifiers. / Hsu, Kuo Wei; Srivastava, Jaideep.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 5476 LNAI 2009. p. 923-932 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5476 LNAI).

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

Hsu, KW & Srivastava, J 2009, Diversity in combinations of heterogeneous classifiers. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 5476 LNAI, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 5476 LNAI, pp. 923-932, 13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2009, Bangkok, Thailand, 27/4/09. https://doi.org/10.1007/978-3-642-01307-2_97
Hsu KW, Srivastava J. Diversity in combinations of heterogeneous classifiers. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 5476 LNAI. 2009. p. 923-932. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-01307-2_97
Hsu, Kuo Wei ; Srivastava, Jaideep. / Diversity in combinations of heterogeneous classifiers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 5476 LNAI 2009. pp. 923-932 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{ff6c975d99be49cbba6024ee0b38b7a1,
title = "Diversity in combinations of heterogeneous classifiers",
abstract = "In this paper, we introduce the use of combinations of heterogeneous classifiers to achieve better diversity. Conducting theoretical and empirical analyses of the diversity of combinations of heterogeneous classifiers, we study the relationship between heterogeneity and diversity. On the one hand, the theoretical analysis serves as a foundation for employing heterogeneous classifiers in Multi-Classifier Systems or ensembles. On the other hand, experimental results provide empirical evidence. We consider synthetic as well as real data sets, utilize classification algorithms that are essentially different, and employ various popular diversity measures for evaluation. Two interesting observations will contribute to the future design of Multi-Classifier Systems and ensemble techniques. First, the diversity among heterogeneous classifiers is higher than that among homogeneous ones, and hence using heterogeneous classifiers to construct classifier combinations would increase the diversity. Second, the heterogeneity primarily results from different classification algorithms rather than the same algorithm with different parameters.",
keywords = "Diversity, Ensemble, Heterogeneity, Multi-classifier system",
author = "Hsu, {Kuo Wei} and Jaideep Srivastava",
year = "2009",
doi = "10.1007/978-3-642-01307-2_97",
language = "English",
isbn = "3642013066",
volume = "5476 LNAI",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "923--932",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Diversity in combinations of heterogeneous classifiers

AU - Hsu, Kuo Wei

AU - Srivastava, Jaideep

PY - 2009

Y1 - 2009

N2 - In this paper, we introduce the use of combinations of heterogeneous classifiers to achieve better diversity. Conducting theoretical and empirical analyses of the diversity of combinations of heterogeneous classifiers, we study the relationship between heterogeneity and diversity. On the one hand, the theoretical analysis serves as a foundation for employing heterogeneous classifiers in Multi-Classifier Systems or ensembles. On the other hand, experimental results provide empirical evidence. We consider synthetic as well as real data sets, utilize classification algorithms that are essentially different, and employ various popular diversity measures for evaluation. Two interesting observations will contribute to the future design of Multi-Classifier Systems and ensemble techniques. First, the diversity among heterogeneous classifiers is higher than that among homogeneous ones, and hence using heterogeneous classifiers to construct classifier combinations would increase the diversity. Second, the heterogeneity primarily results from different classification algorithms rather than the same algorithm with different parameters.

AB - In this paper, we introduce the use of combinations of heterogeneous classifiers to achieve better diversity. Conducting theoretical and empirical analyses of the diversity of combinations of heterogeneous classifiers, we study the relationship between heterogeneity and diversity. On the one hand, the theoretical analysis serves as a foundation for employing heterogeneous classifiers in Multi-Classifier Systems or ensembles. On the other hand, experimental results provide empirical evidence. We consider synthetic as well as real data sets, utilize classification algorithms that are essentially different, and employ various popular diversity measures for evaluation. Two interesting observations will contribute to the future design of Multi-Classifier Systems and ensemble techniques. First, the diversity among heterogeneous classifiers is higher than that among homogeneous ones, and hence using heterogeneous classifiers to construct classifier combinations would increase the diversity. Second, the heterogeneity primarily results from different classification algorithms rather than the same algorithm with different parameters.

KW - Diversity

KW - Ensemble

KW - Heterogeneity

KW - Multi-classifier system

UR - http://www.scopus.com/inward/record.url?scp=67650677731&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=67650677731&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-01307-2_97

DO - 10.1007/978-3-642-01307-2_97

M3 - Conference contribution

SN - 3642013066

SN - 9783642013065

VL - 5476 LNAI

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 923

EP - 932

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -