An improved multi-label classification method based on svm with delicate decision boundary

Benhui Chen, Liangpeng Ma, Takayuki Furuzuki

Research output: Contribution to journalArticle

16 Citations (Scopus)

Abstract

Multi-label classification problem is an extension of traditional multi-class classification problem in which the classes are not mutually exclusive and each sample may belong to several classes simultanrously. Such problems occur in many important applications. Some researches indicate that the performance of classifier can be improved by introducing The information of multi-lahrl training samples into learning procedure effectively. In this paper, we propose a novel method based on SVM with delicate decision boundary. For Thr basic overlapping problem of two lahrls. characteristics of douhlelabel samples arc utilized to obtain Thr range of overlapping sample space decided by two binary SVM classifier separating surfaces. And a bias model with delicate decision boundary is built for samples in overlapping sample space to improve the classification accuracy. Experimental results on the benchmark datasets of Yeast and Scene show that our proposed method improves the classification accuracy efficiently, compared with the basic binary SVM method and some existing well-known methods.

Original languageEnglish
Pages (from-to)1605-1614
Number of pages10
JournalInternational Journal of Innovative Computing, Information and Control
Volume6
Issue number4
Publication statusPublished - 2010 Apr

Fingerprint

Labels
Overlapping
Sample space
Classification Problems
Classifiers
Classifier
Binary
Multi-class Classification
Mutually exclusive
Training Samples
Yeast
Arc of a curve
Benchmark
Experimental Results
Range of data
Class
Model

Keywords

  • Delicate decision boundary
  • Multi-label classification
  • Probabilistic outputs of SVM
  • Support vector machine

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Information Systems
  • Software
  • Theoretical Computer Science

Cite this

An improved multi-label classification method based on svm with delicate decision boundary. / Chen, Benhui; Ma, Liangpeng; Furuzuki, Takayuki.

In: International Journal of Innovative Computing, Information and Control, Vol. 6, No. 4, 04.2010, p. 1605-1614.

Research output: Contribution to journalArticle

@article{1789f1eb039b494eb2ef41b218054828,
title = "An improved multi-label classification method based on svm with delicate decision boundary",
abstract = "Multi-label classification problem is an extension of traditional multi-class classification problem in which the classes are not mutually exclusive and each sample may belong to several classes simultanrously. Such problems occur in many important applications. Some researches indicate that the performance of classifier can be improved by introducing The information of multi-lahrl training samples into learning procedure effectively. In this paper, we propose a novel method based on SVM with delicate decision boundary. For Thr basic overlapping problem of two lahrls. characteristics of douhlelabel samples arc utilized to obtain Thr range of overlapping sample space decided by two binary SVM classifier separating surfaces. And a bias model with delicate decision boundary is built for samples in overlapping sample space to improve the classification accuracy. Experimental results on the benchmark datasets of Yeast and Scene show that our proposed method improves the classification accuracy efficiently, compared with the basic binary SVM method and some existing well-known methods.",
keywords = "Delicate decision boundary, Multi-label classification, Probabilistic outputs of SVM, Support vector machine",
author = "Benhui Chen and Liangpeng Ma and Takayuki Furuzuki",
year = "2010",
month = "4",
language = "English",
volume = "6",
pages = "1605--1614",
journal = "International Journal of Innovative Computing, Information and Control",
issn = "1349-4198",
publisher = "IJICIC Editorial Office",
number = "4",

}

TY - JOUR

T1 - An improved multi-label classification method based on svm with delicate decision boundary

AU - Chen, Benhui

AU - Ma, Liangpeng

AU - Furuzuki, Takayuki

PY - 2010/4

Y1 - 2010/4

N2 - Multi-label classification problem is an extension of traditional multi-class classification problem in which the classes are not mutually exclusive and each sample may belong to several classes simultanrously. Such problems occur in many important applications. Some researches indicate that the performance of classifier can be improved by introducing The information of multi-lahrl training samples into learning procedure effectively. In this paper, we propose a novel method based on SVM with delicate decision boundary. For Thr basic overlapping problem of two lahrls. characteristics of douhlelabel samples arc utilized to obtain Thr range of overlapping sample space decided by two binary SVM classifier separating surfaces. And a bias model with delicate decision boundary is built for samples in overlapping sample space to improve the classification accuracy. Experimental results on the benchmark datasets of Yeast and Scene show that our proposed method improves the classification accuracy efficiently, compared with the basic binary SVM method and some existing well-known methods.

AB - Multi-label classification problem is an extension of traditional multi-class classification problem in which the classes are not mutually exclusive and each sample may belong to several classes simultanrously. Such problems occur in many important applications. Some researches indicate that the performance of classifier can be improved by introducing The information of multi-lahrl training samples into learning procedure effectively. In this paper, we propose a novel method based on SVM with delicate decision boundary. For Thr basic overlapping problem of two lahrls. characteristics of douhlelabel samples arc utilized to obtain Thr range of overlapping sample space decided by two binary SVM classifier separating surfaces. And a bias model with delicate decision boundary is built for samples in overlapping sample space to improve the classification accuracy. Experimental results on the benchmark datasets of Yeast and Scene show that our proposed method improves the classification accuracy efficiently, compared with the basic binary SVM method and some existing well-known methods.

KW - Delicate decision boundary

KW - Multi-label classification

KW - Probabilistic outputs of SVM

KW - Support vector machine

UR - http://www.scopus.com/inward/record.url?scp=77951861159&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77951861159&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:77951861159

VL - 6

SP - 1605

EP - 1614

JO - International Journal of Innovative Computing, Information and Control

JF - International Journal of Innovative Computing, Information and Control

SN - 1349-4198

IS - 4

ER -