EUS SVMs: Ensemble of under-sampled SVMs for data imbalance problems

Pilsung Kang, Sungzoon Cho

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

89 Citations (Scopus)

Abstract

Data imbalance occurs when the number of patterns from a class is much larger than that from the other class. It often degenerates the classification performance. In this paper, we propose an Ensemble of Under-Sampled SVMs or EUS SVMs. We applied the proposed method to two synthetic and six real data sets and we found that it outperformed other methods, especially when the number of patterns belonging to the minority class is very small.

Original languageEnglish
Title of host publicationNeural Information Processing - 13th International Conference, ICONIP 2006, Proceedings
PublisherSpringer Verlag
Pages837-846
Number of pages10
ISBN (Print)3540464794, 9783540464792
Publication statusPublished - 2006 Jan 1
Externally publishedYes
Event13th International Conference on Neural Information Processing, ICONIP 2006 - Hong Kong, China
Duration: 2006 Oct 32006 Oct 6

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4232 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other13th International Conference on Neural Information Processing, ICONIP 2006
CountryChina
CityHong Kong
Period06/10/306/10/6

Fingerprint

Ensemble
Class

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Kang, P., & Cho, S. (2006). EUS SVMs: Ensemble of under-sampled SVMs for data imbalance problems. In Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings (pp. 837-846). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4232 LNCS). Springer Verlag.

EUS SVMs : Ensemble of under-sampled SVMs for data imbalance problems. / Kang, Pilsung; Cho, Sungzoon.

Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings. Springer Verlag, 2006. p. 837-846 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4232 LNCS).

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

Kang, P & Cho, S 2006, EUS SVMs: Ensemble of under-sampled SVMs for data imbalance problems. in Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4232 LNCS, Springer Verlag, pp. 837-846, 13th International Conference on Neural Information Processing, ICONIP 2006, Hong Kong, China, 06/10/3.
Kang P, Cho S. EUS SVMs: Ensemble of under-sampled SVMs for data imbalance problems. In Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings. Springer Verlag. 2006. p. 837-846. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Kang, Pilsung ; Cho, Sungzoon. / EUS SVMs : Ensemble of under-sampled SVMs for data imbalance problems. Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings. Springer Verlag, 2006. pp. 837-846 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{f90d073c42284026ad434e0089f315ef,
title = "EUS SVMs: Ensemble of under-sampled SVMs for data imbalance problems",
abstract = "Data imbalance occurs when the number of patterns from a class is much larger than that from the other class. It often degenerates the classification performance. In this paper, we propose an Ensemble of Under-Sampled SVMs or EUS SVMs. We applied the proposed method to two synthetic and six real data sets and we found that it outperformed other methods, especially when the number of patterns belonging to the minority class is very small.",
author = "Pilsung Kang and Sungzoon Cho",
year = "2006",
month = "1",
day = "1",
language = "English",
isbn = "3540464794",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "837--846",
booktitle = "Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings",

}

TY - GEN

T1 - EUS SVMs

T2 - Ensemble of under-sampled SVMs for data imbalance problems

AU - Kang, Pilsung

AU - Cho, Sungzoon

PY - 2006/1/1

Y1 - 2006/1/1

N2 - Data imbalance occurs when the number of patterns from a class is much larger than that from the other class. It often degenerates the classification performance. In this paper, we propose an Ensemble of Under-Sampled SVMs or EUS SVMs. We applied the proposed method to two synthetic and six real data sets and we found that it outperformed other methods, especially when the number of patterns belonging to the minority class is very small.

AB - Data imbalance occurs when the number of patterns from a class is much larger than that from the other class. It often degenerates the classification performance. In this paper, we propose an Ensemble of Under-Sampled SVMs or EUS SVMs. We applied the proposed method to two synthetic and six real data sets and we found that it outperformed other methods, especially when the number of patterns belonging to the minority class is very small.

UR - http://www.scopus.com/inward/record.url?scp=33750593995&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33750593995&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:33750593995

SN - 3540464794

SN - 9783540464792

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

SP - 837

EP - 846

BT - Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings

PB - Springer Verlag

ER -