Skip to main navigation Skip to search Skip to main content

Single sequential minimal optimization: An improved SVMs training algorithm

  • Harbin Institute of Technology

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

We introduce homogeneous coordinates to represent support vector machines (SVMs) and develop a corresponding training algorithm: single sequential minimal optimization (SSMO). By this simple trick (homogeneous coordinates representation), linear constrains will not appear in quadratic programming (QP) optimization problem. So unlike the most popular used SVM training algorithm sequential minimal optimization (SMO) which solves the QP subproblem containing minimal two Lagrange multipliers, SSMO can analytically update only one Lagrange multiplier at every step. Because of avoiding double loops in heuristically choosing the two Lagrange multipliers in SMO, both CPU time and iterations can be decreased greatly. Experiments on MNIST database, under mild KKT conditions accuracy requirement, shows SSMO can be more than 2 times faster than SMO.

Original languageEnglish
Title of host publication2005 International Conference on Machine Learning and Cybernetics, ICMLC 2005
Pages4360-4364
Number of pages5
StatePublished - 2005
EventInternational Conference on Machine Learning and Cybernetics, ICMLC 2005 - Guangzhou, China
Duration: 18 Aug 200521 Aug 2005

Publication series

Name2005 International Conference on Machine Learning and Cybernetics, ICMLC 2005

Conference

ConferenceInternational Conference on Machine Learning and Cybernetics, ICMLC 2005
Country/TerritoryChina
CityGuangzhou
Period18/08/0521/08/05

Keywords

  • QP
  • Quadratic Programming
  • SMO
  • SSMO
  • SVMs
  • Sequential minimal optimization
  • Single Sequential Minimal Optimization
  • Support Vector Machines

Fingerprint

Dive into the research topics of 'Single sequential minimal optimization: An improved SVMs training algorithm'. Together they form a unique fingerprint.

Cite this