Multi-class cancer classification with OVR-support vector machines selected by Naïve bayes classifier

Jin Hyuk Hong, Sung Bae Cho

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

3 Citations (Scopus)

Abstract

Support vector machines (SVMs), originally designed for binary classification, have been applied for multi-class classification, where an effective fusion scheme is required for combining outputs from them and producing a final result. In this work, we propose a novel method in which the SVMs are generated with the one-vs-rest (OVR) scheme and dynamically organized by the naive Bayes classifiers (NBs). This method might break the ties that frequently occur when working with multi-class classification systems with OVR SVMs. More specifically, we use the Pearson correlation measure to select informative genes and reduce the dimensionality of gene expression profiles when constructing the NBs. The proposed method has been validated on GCM cancer dataset consisting of 14 types of tumors with 16, 063 gene expression levels and produced higher accuracy than other methods.

Original languageEnglish
Title of host publicationNeural Information Processing - 13th International Conference, ICONIP 2006, Proceedings
PublisherSpringer Verlag
Pages155-164
Number of pages10
ISBN (Print)3540464840, 9783540464846
Publication statusPublished - 2006 Jan 1
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)
Volume4234 LNCS - III
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

Bayes Classifier
Cancer Classification
Multi-class Classification
Support vector machines
Support Vector Machine
Classifiers
Gene expression
Naive Bayes Classifier
Pearson Correlation
Gene Expression Profile
Binary Classification
Tumors
Fusion reactions
Tie
Genes
Gene Expression
Dimensionality
Tumor
Cancer
Fusion

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Hong, J. H., & Cho, S. B. (2006). Multi-class cancer classification with OVR-support vector machines selected by Naïve bayes classifier. In Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings (pp. 155-164). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4234 LNCS - III). Springer Verlag.
Hong, Jin Hyuk ; Cho, Sung Bae. / Multi-class cancer classification with OVR-support vector machines selected by Naïve bayes classifier. Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings. Springer Verlag, 2006. pp. 155-164 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{6a085eda79344ad48a76e74ecc3cf07b,
title = "Multi-class cancer classification with OVR-support vector machines selected by Na{\"i}ve bayes classifier",
abstract = "Support vector machines (SVMs), originally designed for binary classification, have been applied for multi-class classification, where an effective fusion scheme is required for combining outputs from them and producing a final result. In this work, we propose a novel method in which the SVMs are generated with the one-vs-rest (OVR) scheme and dynamically organized by the naive Bayes classifiers (NBs). This method might break the ties that frequently occur when working with multi-class classification systems with OVR SVMs. More specifically, we use the Pearson correlation measure to select informative genes and reduce the dimensionality of gene expression profiles when constructing the NBs. The proposed method has been validated on GCM cancer dataset consisting of 14 types of tumors with 16, 063 gene expression levels and produced higher accuracy than other methods.",
author = "Hong, {Jin Hyuk} and Cho, {Sung Bae}",
year = "2006",
month = "1",
day = "1",
language = "English",
isbn = "3540464840",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "155--164",
booktitle = "Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings",
address = "Germany",

}

Hong, JH & Cho, SB 2006, Multi-class cancer classification with OVR-support vector machines selected by Naïve bayes classifier. 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. 4234 LNCS - III, Springer Verlag, pp. 155-164, 13th International Conference on Neural Information Processing, ICONIP 2006, Hong Kong, China, 06/10/3.

Multi-class cancer classification with OVR-support vector machines selected by Naïve bayes classifier. / Hong, Jin Hyuk; Cho, Sung Bae.

Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings. Springer Verlag, 2006. p. 155-164 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4234 LNCS - III).

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

TY - GEN

T1 - Multi-class cancer classification with OVR-support vector machines selected by Naïve bayes classifier

AU - Hong, Jin Hyuk

AU - Cho, Sung Bae

PY - 2006/1/1

Y1 - 2006/1/1

N2 - Support vector machines (SVMs), originally designed for binary classification, have been applied for multi-class classification, where an effective fusion scheme is required for combining outputs from them and producing a final result. In this work, we propose a novel method in which the SVMs are generated with the one-vs-rest (OVR) scheme and dynamically organized by the naive Bayes classifiers (NBs). This method might break the ties that frequently occur when working with multi-class classification systems with OVR SVMs. More specifically, we use the Pearson correlation measure to select informative genes and reduce the dimensionality of gene expression profiles when constructing the NBs. The proposed method has been validated on GCM cancer dataset consisting of 14 types of tumors with 16, 063 gene expression levels and produced higher accuracy than other methods.

AB - Support vector machines (SVMs), originally designed for binary classification, have been applied for multi-class classification, where an effective fusion scheme is required for combining outputs from them and producing a final result. In this work, we propose a novel method in which the SVMs are generated with the one-vs-rest (OVR) scheme and dynamically organized by the naive Bayes classifiers (NBs). This method might break the ties that frequently occur when working with multi-class classification systems with OVR SVMs. More specifically, we use the Pearson correlation measure to select informative genes and reduce the dimensionality of gene expression profiles when constructing the NBs. The proposed method has been validated on GCM cancer dataset consisting of 14 types of tumors with 16, 063 gene expression levels and produced higher accuracy than other methods.

UR - http://www.scopus.com/inward/record.url?scp=33750717515&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33750717515&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:33750717515

SN - 3540464840

SN - 9783540464846

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

SP - 155

EP - 164

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

PB - Springer Verlag

ER -

Hong JH, Cho SB. Multi-class cancer classification with OVR-support vector machines selected by Naïve bayes classifier. In Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings. Springer Verlag. 2006. p. 155-164. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).