Support vector class description (SVCD): Classification in kernel space

Pilsung Kang, Sungzoon Cho

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

We proposed a kernel-based binary classification algorithm, named support vector class description (SVCD), which is an extended version of support vector domain description (SVDD) for one-class classification. SVCD constructs two compact hyperspheres in the feature space such that each hypersphere includes as many instances as possible of one class, while keeping the instances of the other class away from the sphere. By doing this, two linearly non-separable classes in the input space can be well distinguished in the feature space. In order to verify the classification performance and exploit the properties of SVCD, we conducted experiments on actual classification data sets and analyzed the results. Compared with other popular kernel-based classification algorithms, such as support vector machine (SVM) and kernel Fisher discriminant analysis (KFD), SVCD gave better classification performances in terms of both the area under the receiving operator curve (AUROC) and the balanced correction rate (BCR). In addition, SVCD was found to be capable of finding moderate sparse solutions with little parameter sensitivity.

Original languageEnglish
Pages (from-to)351-364
Number of pages14
JournalIntelligent Data Analysis
Volume16
Issue number3
DOIs
Publication statusPublished - 2012 May 28
Externally publishedYes

Fingerprint

Support Vector
kernel
Hypersphere
Classification Algorithm
Feature Space
One-class Classification
Discriminant analysis
Fisher Discriminant Analysis
Parameter Sensitivity
Binary Classification
Support vector machines
Data Classification
Class
Nonseparable
Support Vector Machine
Linearly
Verify
Curve
Operator
Experiments

Keywords

  • classification
  • kernel learning
  • Support vector learning

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Vision and Pattern Recognition
  • Artificial Intelligence

Cite this

Support vector class description (SVCD) : Classification in kernel space. / Kang, Pilsung; Cho, Sungzoon.

In: Intelligent Data Analysis, Vol. 16, No. 3, 28.05.2012, p. 351-364.

Research output: Contribution to journalArticle

@article{4786c82b2fa84277945698a9019de29c,
title = "Support vector class description (SVCD): Classification in kernel space",
abstract = "We proposed a kernel-based binary classification algorithm, named support vector class description (SVCD), which is an extended version of support vector domain description (SVDD) for one-class classification. SVCD constructs two compact hyperspheres in the feature space such that each hypersphere includes as many instances as possible of one class, while keeping the instances of the other class away from the sphere. By doing this, two linearly non-separable classes in the input space can be well distinguished in the feature space. In order to verify the classification performance and exploit the properties of SVCD, we conducted experiments on actual classification data sets and analyzed the results. Compared with other popular kernel-based classification algorithms, such as support vector machine (SVM) and kernel Fisher discriminant analysis (KFD), SVCD gave better classification performances in terms of both the area under the receiving operator curve (AUROC) and the balanced correction rate (BCR). In addition, SVCD was found to be capable of finding moderate sparse solutions with little parameter sensitivity.",
keywords = "classification, kernel learning, Support vector learning",
author = "Pilsung Kang and Sungzoon Cho",
year = "2012",
month = "5",
day = "28",
doi = "10.3233/IDA-2012-0528",
language = "English",
volume = "16",
pages = "351--364",
journal = "Intelligent Data Analysis",
issn = "1088-467X",
publisher = "IOS Press",
number = "3",

}

TY - JOUR

T1 - Support vector class description (SVCD)

T2 - Classification in kernel space

AU - Kang, Pilsung

AU - Cho, Sungzoon

PY - 2012/5/28

Y1 - 2012/5/28

N2 - We proposed a kernel-based binary classification algorithm, named support vector class description (SVCD), which is an extended version of support vector domain description (SVDD) for one-class classification. SVCD constructs two compact hyperspheres in the feature space such that each hypersphere includes as many instances as possible of one class, while keeping the instances of the other class away from the sphere. By doing this, two linearly non-separable classes in the input space can be well distinguished in the feature space. In order to verify the classification performance and exploit the properties of SVCD, we conducted experiments on actual classification data sets and analyzed the results. Compared with other popular kernel-based classification algorithms, such as support vector machine (SVM) and kernel Fisher discriminant analysis (KFD), SVCD gave better classification performances in terms of both the area under the receiving operator curve (AUROC) and the balanced correction rate (BCR). In addition, SVCD was found to be capable of finding moderate sparse solutions with little parameter sensitivity.

AB - We proposed a kernel-based binary classification algorithm, named support vector class description (SVCD), which is an extended version of support vector domain description (SVDD) for one-class classification. SVCD constructs two compact hyperspheres in the feature space such that each hypersphere includes as many instances as possible of one class, while keeping the instances of the other class away from the sphere. By doing this, two linearly non-separable classes in the input space can be well distinguished in the feature space. In order to verify the classification performance and exploit the properties of SVCD, we conducted experiments on actual classification data sets and analyzed the results. Compared with other popular kernel-based classification algorithms, such as support vector machine (SVM) and kernel Fisher discriminant analysis (KFD), SVCD gave better classification performances in terms of both the area under the receiving operator curve (AUROC) and the balanced correction rate (BCR). In addition, SVCD was found to be capable of finding moderate sparse solutions with little parameter sensitivity.

KW - classification

KW - kernel learning

KW - Support vector learning

UR - http://www.scopus.com/inward/record.url?scp=84861384614&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84861384614&partnerID=8YFLogxK

U2 - 10.3233/IDA-2012-0528

DO - 10.3233/IDA-2012-0528

M3 - Article

AN - SCOPUS:84861384614

VL - 16

SP - 351

EP - 364

JO - Intelligent Data Analysis

JF - Intelligent Data Analysis

SN - 1088-467X

IS - 3

ER -