English
 
User Manual Privacy Policy Disclaimer Contact us
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT
  Single-class Support Vector Machines

Schölkopf, B., Williamson, R., Smola, A., & Shawe-Taylor, J. (1999). Single-class Support Vector Machines. Poster presented at Dagstuhl-Seminar 99121: Unsupervised Learning, Dagstuhl, Germany.

Item is

Basic

show hide
Item Permalink: http://hdl.handle.net/11858/00-001M-0000-0013-E79A-2 Version Permalink: http://hdl.handle.net/21.11116/0000-0005-C58C-F
Genre: Poster

Files

show Files

Locators

show
hide
Description:
-

Creators

show
hide
 Creators:
Schölkopf, B1, Author              
Williamson, R, Author
Smola, AJ, Author              
Shawe-Taylor, J, Author
Affiliations:
1External Organizations, ou_persistent22              

Content

show
hide
Free keywords: -
 Abstract: Suppose you are given some dataset drawn from an underlying probability dis-tributionPand you want to estimate a subsetSof input space such that theprobability that a test point drawn fromPlies outside ofSis bounded by somea priori specified 0< ν≤1.We propose an algorithm to deal with this problem by trying toestimate afunctionfwhich is positive onSand negative on the complement ofS. Thefunctional form offis given by a kernel expansion in terms of a potentially smallsubset of the training data; it is regularized by controlling the length of the weightvector in an associated feature space.We can prove thatνupper bounds the fraction of outliers (training points outsideofS) and lower bounds the fraction of support vectors. Asymptotically, undersome mild condition onP, both become equalities.The algorithm is a natural extension of the support vector algorithm to the case of unlabelled data.

Details

show
hide
Language(s):
 Dates: 1999-03
 Publication Status: Published in print
 Pages: -
 Publishing info: -
 Table of Contents: -
 Rev. Type: -
 Identifiers: BibTex Citekey: 1828
 Degree: -

Event

show
hide
Title: Dagstuhl-Seminar 99121: Unsupervised Learning
Place of Event: Dagstuhl, Germany
Start-/End Date: 1999-03-21 - 1999-03-26

Legal Case

show

Project information

show

Source 1

show
hide
Title: Dagstuhl Reports
Source Genre: Journal
 Creator(s):
Affiliations:
Publ. Info: Wadern : Schloss Dagstuhl, Leibniz-Zentrum für Informatik
Pages: - Volume / Issue: 235 Sequence Number: 22 Start / End Page: 19 - 20 Identifier: ISSN: 2192-5283
CoNE: https://pure.mpg.de/cone/journals/resource/21925283