Communications for Statistical Applications and Methods

The Korean Statistical Society (한국통계학회)
권호 표지
DOI QR코드

DOI QR Code

A Study on Support Vectors of Least Squares Support Vector Machine

  • Seok, Kyungha (Department of Data Science, Institute of Basic Sciences, Inje University) ;
  • Cho, Daehyun (Department of Data Science, Institute of Basic Sciences, Inje University)
  • Published : 2003.12.01
Download PDF
⟨ Previous Next ⟩

Abstract

LS-SVM(Least-Squares Support Vector Machine) has been used as a promising method for regression as well as classification. Suykens et al.(2000) used only the magnitude of residuals to obtain SVs(Support Vectors). Suykens' method behaves well for homogeneous model. But in a heteroscedastic model, the method shows a poor behavior. The present paper proposes a new method to get SVs. The proposed method uses the variance of noise as well as the magnitude of residuals to obtain support vectors. Through the simulation study we justified excellence of our proposed method.

Keywords

References

  1. Neural networks for pattern recognition Bishop,C.M.
  2. Learning from data: concept, theory and method Cherkassky,V.;Mulier,F.
  3. An Introduction to Support Vector Machines Cristianini,N.;Shawe-Taylor,J.
  4. Neural Networks: a Comprehensive Foundation Haykin,S.
  5. Neural Processing Letters v.9 no.3 Least squares support vector machine classifiers Suykens,J.A.K.;Vandewalle,J. https://doi.org/10.1023/A:1018628609742
  6. IEEE International Symposium on Circuits and Systems (ISCAS 20000 Sparce approximation using least squares support vector machines Suykens,J.A.K.;Lukas,L.;Vandewalle,J.
  7. Neural Networks v.14 no.1 Optimal control by least squares support vector machines Suykens,J.A.K.;Vandewalle,J.;De Moor B. https://doi.org/10.1016/S0893-6080(00)00077-0
  8. Internal Report 00-37, ESAT-SISTA, K.U. Leuven Benchmarking least squares support vector machine classifiers Van Gestel T.;Suykens,J.A.K.;Baesens,B.;Viaene,S.;Vanthienen,J.;Dedene,G.;De Moor B.
  9. Statistical learning theory Vapnik,V.