Graph-based semi-supervised support vector data description for novelty detection

Phuong Duong, Van Nguyen, Mi Dinh, Trung Le, Dat TRAN, Wanli MA

Research output: A Conference proceeding or a Chapter in BookConference contribution

6 Citations (Scopus)
3 Downloads (Pure)

Abstract

Support Vector Data Description (SVDD) is a well-known supervised learning method for novelty detection purpose. For its classification task, SVDD requires a fully-labeled dataset. Nonetheless, contemporary datasets always consist of a collection of labeled data samples jointly a much larger collection of unlabeled ones. This fact impedes the usage of SVDD in the real-world problems. In this paper, we propose to utilize the information implicated in a spectral graph to leverage SVDD in the context of semi-supervised learning. The theory and experiment evidence that the proposed method is able to efficiently employ the information carried in the spectral graph to not only enhance the generalization ability of SVDD but also enforce the cluster assumption which is crucial for a semi-supervised learning method
Original languageEnglish
Title of host publication2015 International joint conference on neural networks (IJCNN 2015)
EditorsAmir Hussain
Place of PublicationUSA
PublisherIEEE, Institute of Electrical and Electronics Engineers
Pages1985-1990
Number of pages6
Volume1
ISBN (Electronic)9781479919604
ISBN (Print)9781479919611
DOIs
Publication statusPublished - 2015
EventInternational Joint Conference on Neural Networks IJCNN 2015 - Killarney, Ireland, Killarney, Ireland
Duration: 12 Jul 201517 Jul 2015

Publication series

NameProceedings of the International Joint Conference on Neural Networks
Volume2015-September

Conference

ConferenceInternational Joint Conference on Neural Networks IJCNN 2015
Abbreviated titleIJCNN 2015
CountryIreland
CityKillarney
Period12/07/1517/07/15

Fingerprint

Data description
Supervised learning
Experiments

Cite this

Duong, P., Nguyen, V., Dinh, M., Le, T., TRAN, D., & MA, W. (2015). Graph-based semi-supervised support vector data description for novelty detection. In A. Hussain (Ed.), 2015 International joint conference on neural networks (IJCNN 2015) (Vol. 1, pp. 1985-1990). [7280565] (Proceedings of the International Joint Conference on Neural Networks; Vol. 2015-September). USA: IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/ijcnn.2015.7280565
Duong, Phuong ; Nguyen, Van ; Dinh, Mi ; Le, Trung ; TRAN, Dat ; MA, Wanli. / Graph-based semi-supervised support vector data description for novelty detection. 2015 International joint conference on neural networks (IJCNN 2015). editor / Amir Hussain. Vol. 1 USA : IEEE, Institute of Electrical and Electronics Engineers, 2015. pp. 1985-1990 (Proceedings of the International Joint Conference on Neural Networks).
@inproceedings{13e09370452d4c6b958c8d2d0d040e39,
title = "Graph-based semi-supervised support vector data description for novelty detection",
abstract = "Support Vector Data Description (SVDD) is a well-known supervised learning method for novelty detection purpose. For its classification task, SVDD requires a fully-labeled dataset. Nonetheless, contemporary datasets always consist of a collection of labeled data samples jointly a much larger collection of unlabeled ones. This fact impedes the usage of SVDD in the real-world problems. In this paper, we propose to utilize the information implicated in a spectral graph to leverage SVDD in the context of semi-supervised learning. The theory and experiment evidence that the proposed method is able to efficiently employ the information carried in the spectral graph to not only enhance the generalization ability of SVDD but also enforce the cluster assumption which is crucial for a semi-supervised learning method",
keywords = "Kernel method, novelty detection, one-class classification, semi-supervised learning, subspace learning",
author = "Phuong Duong and Van Nguyen and Mi Dinh and Trung Le and Dat TRAN and Wanli MA",
year = "2015",
doi = "10.1109/ijcnn.2015.7280565",
language = "English",
isbn = "9781479919611",
volume = "1",
series = "Proceedings of the International Joint Conference on Neural Networks",
publisher = "IEEE, Institute of Electrical and Electronics Engineers",
pages = "1985--1990",
editor = "Amir Hussain",
booktitle = "2015 International joint conference on neural networks (IJCNN 2015)",
address = "United States",

}

Duong, P, Nguyen, V, Dinh, M, Le, T, TRAN, D & MA, W 2015, Graph-based semi-supervised support vector data description for novelty detection. in A Hussain (ed.), 2015 International joint conference on neural networks (IJCNN 2015). vol. 1, 7280565, Proceedings of the International Joint Conference on Neural Networks, vol. 2015-September, IEEE, Institute of Electrical and Electronics Engineers, USA, pp. 1985-1990, International Joint Conference on Neural Networks IJCNN 2015, Killarney, Ireland, 12/07/15. https://doi.org/10.1109/ijcnn.2015.7280565

Graph-based semi-supervised support vector data description for novelty detection. / Duong, Phuong; Nguyen, Van; Dinh, Mi; Le, Trung; TRAN, Dat; MA, Wanli.

2015 International joint conference on neural networks (IJCNN 2015). ed. / Amir Hussain. Vol. 1 USA : IEEE, Institute of Electrical and Electronics Engineers, 2015. p. 1985-1990 7280565 (Proceedings of the International Joint Conference on Neural Networks; Vol. 2015-September).

Research output: A Conference proceeding or a Chapter in BookConference contribution

TY - GEN

T1 - Graph-based semi-supervised support vector data description for novelty detection

AU - Duong, Phuong

AU - Nguyen, Van

AU - Dinh, Mi

AU - Le, Trung

AU - TRAN, Dat

AU - MA, Wanli

PY - 2015

Y1 - 2015

N2 - Support Vector Data Description (SVDD) is a well-known supervised learning method for novelty detection purpose. For its classification task, SVDD requires a fully-labeled dataset. Nonetheless, contemporary datasets always consist of a collection of labeled data samples jointly a much larger collection of unlabeled ones. This fact impedes the usage of SVDD in the real-world problems. In this paper, we propose to utilize the information implicated in a spectral graph to leverage SVDD in the context of semi-supervised learning. The theory and experiment evidence that the proposed method is able to efficiently employ the information carried in the spectral graph to not only enhance the generalization ability of SVDD but also enforce the cluster assumption which is crucial for a semi-supervised learning method

AB - Support Vector Data Description (SVDD) is a well-known supervised learning method for novelty detection purpose. For its classification task, SVDD requires a fully-labeled dataset. Nonetheless, contemporary datasets always consist of a collection of labeled data samples jointly a much larger collection of unlabeled ones. This fact impedes the usage of SVDD in the real-world problems. In this paper, we propose to utilize the information implicated in a spectral graph to leverage SVDD in the context of semi-supervised learning. The theory and experiment evidence that the proposed method is able to efficiently employ the information carried in the spectral graph to not only enhance the generalization ability of SVDD but also enforce the cluster assumption which is crucial for a semi-supervised learning method

KW - Kernel method

KW - novelty detection

KW - one-class classification

KW - semi-supervised learning

KW - subspace learning

UR - http://www.scopus.com/inward/record.url?scp=84951185209&partnerID=8YFLogxK

U2 - 10.1109/ijcnn.2015.7280565

DO - 10.1109/ijcnn.2015.7280565

M3 - Conference contribution

SN - 9781479919611

VL - 1

T3 - Proceedings of the International Joint Conference on Neural Networks

SP - 1985

EP - 1990

BT - 2015 International joint conference on neural networks (IJCNN 2015)

A2 - Hussain, Amir

PB - IEEE, Institute of Electrical and Electronics Engineers

CY - USA

ER -

Duong P, Nguyen V, Dinh M, Le T, TRAN D, MA W. Graph-based semi-supervised support vector data description for novelty detection. In Hussain A, editor, 2015 International joint conference on neural networks (IJCNN 2015). Vol. 1. USA: IEEE, Institute of Electrical and Electronics Engineers. 2015. p. 1985-1990. 7280565. (Proceedings of the International Joint Conference on Neural Networks). https://doi.org/10.1109/ijcnn.2015.7280565