Convergence of a generalized SMO algorithm for SVM classifier design (2000) [31 citations — 1 self]
by S. S. Keerthi, E. G. Gilbert
Machine Learning
http://guppy.mpe.nus.edu.sg/~mpessk/svm/conv_ml.ps.gz
Add To MetaCart
Popular Tags
No tags have been applied to this document.
BibTeX | Add To MetaCart
@INPROCEEDINGS{Keerthi00convergenceof,author = {S. S. Keerthi and E. G. Gilbert},
title = {Convergence of a generalized SMO algorithm for SVM classifier design},
booktitle = {Machine Learning},
year = {2000},
pages = {351--360}
}
Years of Citing Articles
Abstract:
Convergence of a generalized version of the modied SMO algorithms given by Keerthi et.al. for SVM classier design is proved. The convergence results are also extended to modi ed SMO algorithms for solving -SVM classier problems. 1 Introduction. Platt's Sequential Minimization Algorithm (SMO) (Platt, 1998) is a simple and ecient algorithm for solving the quadratic programming problem arising in support vector machines. Recently Keerthi et.al. (Keerthi et.al, 1999) pointed out a problem caused by the way SMO maintains and updates a single threshold value and suggested two modied versions of SMO
Citations
| 111 | Improvements to Platt’s SMO Algorithm for SVM Classifier Design – Keerthi, Shevade, et al. |
| 70 | On the convergence of the decomposition method for support vector machines – Lin |
