A probabilistic multi-class strategy of one-vs.-rest support vector machines for cancer classification

Jin Hyuk Hong, Sung-Bae Cho

Research output: Contribution to journalArticle

35 Citations (Scopus)

Abstract

Support vector machines (SVMs), originally designed for binary classification, have been applied for multi-class classification with effective decomposition and reconstruction schemes. Decomposition schemes such as one-vs.-rest (OVR) and pair-wise partition a dataset into several subsets of two classes so as to produce multiple outputs that should be combined. Majority voting or winner-takes-all is a representative reconstruction scheme to combine those outputs, but it often causes some problems to consider tie-breaks and tune the weights of individual classifiers. In this paper, we propose a novel method in which SVMs are generated with the OVR scheme and probabilistically ordered by using the naïve Bayes classifiers (NBs). This method is able to break ties that frequently occur when working with multi-class classification systems with OVR SVMs. More specifically, we use the Pearson correlation to select informative genes and reduce the dimensionality of gene expression profiles when constructing the NBs. The proposed method has been validated on several popular multi-class cancer datasets and produced higher accuracy than conventional methods.

Original languageEnglish
Pages (from-to)3275-3281
Number of pages7
JournalNeurocomputing
Volume71
Issue number16-18
DOIs
Publication statusPublished - 2008 Oct 1

Fingerprint

Support vector machines
Classifiers
Decomposition
Neoplasms
Gene expression
Politics
Genes
Transcriptome
Weights and Measures
Support Vector Machine
Datasets

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Cognitive Neuroscience
  • Artificial Intelligence

Cite this

@article{22d370476290427cbab37ef86cc35784,
title = "A probabilistic multi-class strategy of one-vs.-rest support vector machines for cancer classification",
abstract = "Support vector machines (SVMs), originally designed for binary classification, have been applied for multi-class classification with effective decomposition and reconstruction schemes. Decomposition schemes such as one-vs.-rest (OVR) and pair-wise partition a dataset into several subsets of two classes so as to produce multiple outputs that should be combined. Majority voting or winner-takes-all is a representative reconstruction scheme to combine those outputs, but it often causes some problems to consider tie-breaks and tune the weights of individual classifiers. In this paper, we propose a novel method in which SVMs are generated with the OVR scheme and probabilistically ordered by using the na{\"i}ve Bayes classifiers (NBs). This method is able to break ties that frequently occur when working with multi-class classification systems with OVR SVMs. More specifically, we use the Pearson correlation to select informative genes and reduce the dimensionality of gene expression profiles when constructing the NBs. The proposed method has been validated on several popular multi-class cancer datasets and produced higher accuracy than conventional methods.",
author = "Hong, {Jin Hyuk} and Sung-Bae Cho",
year = "2008",
month = "10",
day = "1",
doi = "10.1016/j.neucom.2008.04.033",
language = "English",
volume = "71",
pages = "3275--3281",
journal = "Neurocomputing",
issn = "0925-2312",
publisher = "Elsevier",
number = "16-18",

}

A probabilistic multi-class strategy of one-vs.-rest support vector machines for cancer classification. / Hong, Jin Hyuk; Cho, Sung-Bae.

In: Neurocomputing, Vol. 71, No. 16-18, 01.10.2008, p. 3275-3281.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A probabilistic multi-class strategy of one-vs.-rest support vector machines for cancer classification

AU - Hong, Jin Hyuk

AU - Cho, Sung-Bae

PY - 2008/10/1

Y1 - 2008/10/1

N2 - Support vector machines (SVMs), originally designed for binary classification, have been applied for multi-class classification with effective decomposition and reconstruction schemes. Decomposition schemes such as one-vs.-rest (OVR) and pair-wise partition a dataset into several subsets of two classes so as to produce multiple outputs that should be combined. Majority voting or winner-takes-all is a representative reconstruction scheme to combine those outputs, but it often causes some problems to consider tie-breaks and tune the weights of individual classifiers. In this paper, we propose a novel method in which SVMs are generated with the OVR scheme and probabilistically ordered by using the naïve Bayes classifiers (NBs). This method is able to break ties that frequently occur when working with multi-class classification systems with OVR SVMs. More specifically, we use the Pearson correlation to select informative genes and reduce the dimensionality of gene expression profiles when constructing the NBs. The proposed method has been validated on several popular multi-class cancer datasets and produced higher accuracy than conventional methods.

AB - Support vector machines (SVMs), originally designed for binary classification, have been applied for multi-class classification with effective decomposition and reconstruction schemes. Decomposition schemes such as one-vs.-rest (OVR) and pair-wise partition a dataset into several subsets of two classes so as to produce multiple outputs that should be combined. Majority voting or winner-takes-all is a representative reconstruction scheme to combine those outputs, but it often causes some problems to consider tie-breaks and tune the weights of individual classifiers. In this paper, we propose a novel method in which SVMs are generated with the OVR scheme and probabilistically ordered by using the naïve Bayes classifiers (NBs). This method is able to break ties that frequently occur when working with multi-class classification systems with OVR SVMs. More specifically, we use the Pearson correlation to select informative genes and reduce the dimensionality of gene expression profiles when constructing the NBs. The proposed method has been validated on several popular multi-class cancer datasets and produced higher accuracy than conventional methods.

UR - http://www.scopus.com/inward/record.url?scp=56549128359&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=56549128359&partnerID=8YFLogxK

U2 - 10.1016/j.neucom.2008.04.033

DO - 10.1016/j.neucom.2008.04.033

M3 - Article

AN - SCOPUS:56549128359

VL - 71

SP - 3275

EP - 3281

JO - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

IS - 16-18

ER -