Improving SVM based multi-label classification by using label relationship

Di Fu, Bo Zhou, Takayuki Furuzuki

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

2 Citations (Scopus)

Abstract

This paper proposes an improved SVM based multi-label classification method by using relationship among labels. Following a traditional multi-label solution, binary relevance (BR) method is first used to decompose the multi-label classification problem into multiple binary classification sub-problems, each of which is solved by an SVM classifier. By using Platt's sigmoid technique, each SVM classifier gives probability output for the following correction. A probability model is introduced to estimate the relationship among labels. The extracted label relationship is then applied to correct the outputs of SVM classifiers, in which a dynamic weight strategy is further introduced. Numerical experiments on widely used benchmark datasets show that the proposed method can improve the accuracy of multi-label classification when compared with traditional BR method and some other conventional multi-label classification methods.

Original languageEnglish
Title of host publicationProceedings of the International Joint Conference on Neural Networks
PublisherInstitute of Electrical and Electronics Engineers Inc.
Volume2015-September
ISBN (Print)9781479919604, 9781479919604, 9781479919604, 9781479919604
DOIs
Publication statusPublished - 2015 Sep 28
EventInternational Joint Conference on Neural Networks, IJCNN 2015 - Killarney, Ireland
Duration: 2015 Jul 122015 Jul 17

Other

OtherInternational Joint Conference on Neural Networks, IJCNN 2015
CountryIreland
CityKillarney
Period15/7/1215/7/17

Fingerprint

Labels
Classifiers
Experiments

Keywords

  • Accuracy
  • Support vector machines

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence

Cite this

Fu, D., Zhou, B., & Furuzuki, T. (2015). Improving SVM based multi-label classification by using label relationship. In Proceedings of the International Joint Conference on Neural Networks (Vol. 2015-September). [7280497] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IJCNN.2015.7280497

Improving SVM based multi-label classification by using label relationship. / Fu, Di; Zhou, Bo; Furuzuki, Takayuki.

Proceedings of the International Joint Conference on Neural Networks. Vol. 2015-September Institute of Electrical and Electronics Engineers Inc., 2015. 7280497.

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

Fu, D, Zhou, B & Furuzuki, T 2015, Improving SVM based multi-label classification by using label relationship. in Proceedings of the International Joint Conference on Neural Networks. vol. 2015-September, 7280497, Institute of Electrical and Electronics Engineers Inc., International Joint Conference on Neural Networks, IJCNN 2015, Killarney, Ireland, 15/7/12. https://doi.org/10.1109/IJCNN.2015.7280497
Fu D, Zhou B, Furuzuki T. Improving SVM based multi-label classification by using label relationship. In Proceedings of the International Joint Conference on Neural Networks. Vol. 2015-September. Institute of Electrical and Electronics Engineers Inc. 2015. 7280497 https://doi.org/10.1109/IJCNN.2015.7280497
Fu, Di ; Zhou, Bo ; Furuzuki, Takayuki. / Improving SVM based multi-label classification by using label relationship. Proceedings of the International Joint Conference on Neural Networks. Vol. 2015-September Institute of Electrical and Electronics Engineers Inc., 2015.
@inproceedings{203f9b2694784058b74d38c24c1b30ec,
title = "Improving SVM based multi-label classification by using label relationship",
abstract = "This paper proposes an improved SVM based multi-label classification method by using relationship among labels. Following a traditional multi-label solution, binary relevance (BR) method is first used to decompose the multi-label classification problem into multiple binary classification sub-problems, each of which is solved by an SVM classifier. By using Platt's sigmoid technique, each SVM classifier gives probability output for the following correction. A probability model is introduced to estimate the relationship among labels. The extracted label relationship is then applied to correct the outputs of SVM classifiers, in which a dynamic weight strategy is further introduced. Numerical experiments on widely used benchmark datasets show that the proposed method can improve the accuracy of multi-label classification when compared with traditional BR method and some other conventional multi-label classification methods.",
keywords = "Accuracy, Support vector machines",
author = "Di Fu and Bo Zhou and Takayuki Furuzuki",
year = "2015",
month = "9",
day = "28",
doi = "10.1109/IJCNN.2015.7280497",
language = "English",
isbn = "9781479919604",
volume = "2015-September",
booktitle = "Proceedings of the International Joint Conference on Neural Networks",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Improving SVM based multi-label classification by using label relationship

AU - Fu, Di

AU - Zhou, Bo

AU - Furuzuki, Takayuki

PY - 2015/9/28

Y1 - 2015/9/28

N2 - This paper proposes an improved SVM based multi-label classification method by using relationship among labels. Following a traditional multi-label solution, binary relevance (BR) method is first used to decompose the multi-label classification problem into multiple binary classification sub-problems, each of which is solved by an SVM classifier. By using Platt's sigmoid technique, each SVM classifier gives probability output for the following correction. A probability model is introduced to estimate the relationship among labels. The extracted label relationship is then applied to correct the outputs of SVM classifiers, in which a dynamic weight strategy is further introduced. Numerical experiments on widely used benchmark datasets show that the proposed method can improve the accuracy of multi-label classification when compared with traditional BR method and some other conventional multi-label classification methods.

AB - This paper proposes an improved SVM based multi-label classification method by using relationship among labels. Following a traditional multi-label solution, binary relevance (BR) method is first used to decompose the multi-label classification problem into multiple binary classification sub-problems, each of which is solved by an SVM classifier. By using Platt's sigmoid technique, each SVM classifier gives probability output for the following correction. A probability model is introduced to estimate the relationship among labels. The extracted label relationship is then applied to correct the outputs of SVM classifiers, in which a dynamic weight strategy is further introduced. Numerical experiments on widely used benchmark datasets show that the proposed method can improve the accuracy of multi-label classification when compared with traditional BR method and some other conventional multi-label classification methods.

KW - Accuracy

KW - Support vector machines

UR - http://www.scopus.com/inward/record.url?scp=84951191702&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84951191702&partnerID=8YFLogxK

U2 - 10.1109/IJCNN.2015.7280497

DO - 10.1109/IJCNN.2015.7280497

M3 - Conference contribution

AN - SCOPUS:84951191702

SN - 9781479919604

SN - 9781479919604

SN - 9781479919604

SN - 9781479919604

VL - 2015-September

BT - Proceedings of the International Joint Conference on Neural Networks

PB - Institute of Electrical and Electronics Engineers Inc.

ER -