Mixture of Support Vector Data Descriptions

Vinh Lai, Duy Nguyen, Khanh Nguyen, Trung Le

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

5 Citations (Scopus)

Abstract

We present Mixture of Support Vector Data Descriptions (mSVDD) for one-class classification or novelty detection. A mixture of optimal hyperspheres is automatically discovered to describe data. The model consists of two parts: log likelihood to control the fit of data to model (empirical risk) and regularization quantizer to control the generalization ability of model (general risk). Expectation Maximization (EM) principle is employed to train the model. We demonstrate the advantage of the proposed model: if learning mSVDD in the input space, it simulates learning single hypersphere in the feature space and the accuracy is thus comparable but the training time is significantly shorter.

Original languageEnglish
Title of host publicationProceedings of 2015 2nd National Foundation for Science and Technology Development Conference on Information and Computer Science
EditorsNguyen Quoc Dinh, Tran Xuan Tu
Place of PublicationPiscataway NJ USA
PublisherIEEE, Institute of Electrical and Electronics Engineers
Pages135-140
Number of pages6
ISBN (Electronic)9781467366403, 9781467366380
ISBN (Print)9781467366397
DOIs
Publication statusPublished - 2015
Externally publishedYes
EventNational Foundation for Science and Technology Development Conference on Information and Computer Science 2015 - Ho Chi Minh, Vietnam
Duration: 16 Sept 201518 Sept 2015
Conference number: 1st
https://edas.info/web/nics14/index.html
https://web.archive.org/web/20150807190107/http://nafosted-nics.org/

Conference

ConferenceNational Foundation for Science and Technology Development Conference on Information and Computer Science 2015
Abbreviated titleNICS 2015
Country/TerritoryVietnam
CityHo Chi Minh
Period16/09/1518/09/15
Internet address

Keywords

  • kernel method
  • mixture model
  • Mixture of experts
  • one-class classification

Cite this