Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Global convergence of SMO algorithm for support vector regression.

Norikazu Takahashi1, Jun Guo, Tetsuo Nishi

  • 1Department of Computer Science and Communication Engineering, Kyushu University, Fukuoka 819-0395, Japan. norikazu@csce.kyushu-u.ac.jp

IEEE Transactions on Neural Networks
|June 11, 2008
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Complete genome sequence of Streptomyces vietnamensis GIMV4.0001 T, a genetically manipulable producer of the benzoisochromanequinone antibiotic granaticin.

Journal of biotechnology·2015
Same author

Controlling and self assembling of monodisperse platinum nanocubes as efficient methanol oxidation electrocatalysts.

Chemical communications (Cambridge, England)·2015
Same author

Pd-nanoparticle-supported, PDDA-functionalized graphene as a promising catalyst for alcohol oxidation.

Chemistry, an Asian journal·2015
Same author

[Efficacy and safety of Longjintonglin Capsule for the treatment of type III prostatitis].

Zhonghua nan ke xue = National journal of andrology·2015
Same author

Prognostic factors for conjunctival melanoma: a study in ethnic Chinese patients.

The British journal of ophthalmology·2015
Same author

Twist1 correlates with poor differentiation and progression in gastric adenocarcinoma via elevation of FGFR2 expression.

World journal of gastroenterology·2015
Same journal

Universal perceptron and DNA-like learning algorithm for binary neural networks: LSBF and PBF implementations.

IEEE transactions on neural networks·2013
Same journal

Guest editorial: special section on white box nonlinear prediction models.

IEEE transactions on neural networks·2011
Same journal

Data-based fault-tolerant control of high-speed trains with traction/braking notch nonlinearities and actuator failures.

IEEE transactions on neural networks·2011
Same journal

Guest editorial: special section on data-based control, modeling, and optimization.

IEEE transactions on neural networks·2011
Same journal

Neural network-based multiple robot simultaneous localization and mapping.

IEEE transactions on neural networks·2011
Same journal

Data-driven model-free adaptive control for a class of MIMO nonlinear discrete-time systems.

IEEE transactions on neural networks·2011
See all related articles

This study proves the global convergence of the sequential minimal optimization (SMO) algorithm for support vector regression (SVR). Efficient implementation techniques are presented, ensuring the SMO algorithm finds optimal solutions in finite iterations.

Area of Science:

  • Machine Learning
  • Optimization Algorithms

Background:

  • Support Vector Regression (SVR) is a powerful technique for regression analysis.
  • The Sequential Minimal Optimization (SMO) algorithm is commonly used for SVR training.
  • Ensuring the convergence of SMO for SVR is crucial for reliable model performance.

Purpose of the Study:

  • To analyze the global convergence properties of the SMO algorithm applied to Support Vector Regression (SVR).
  • To demonstrate that SMO, under specific update conditions, guarantees convergence to an optimal solution in a finite number of steps.
  • To present and experimentally evaluate efficient implementation techniques for the SMO algorithm in SVR.

Main Methods:

  • Formulating SVR as a convex quadratic programming (QP) problem with 'l' pairs of variables.

Related Experiment Videos

  • Proving convergence by analyzing the update steps involving pairs of variables violating optimality conditions.
  • Developing and implementing efficient SMO techniques for practical SVR applications.

    • Main Results:

    • Theoretical proof of global convergence for the SMO algorithm in SVR.
    • Demonstration that specific update strategies ensure finite iteration convergence to an optimal solution.
    • Experimental comparison of proposed efficient SMO implementations against existing SMO algorithms.

    Conclusions:

    • The SMO algorithm is guaranteed to converge globally for SVR under the proposed conditions.
    • Efficient implementation techniques enhance the practical applicability and performance of SMO for SVR.
    • This work provides a theoretical foundation and practical improvements for SMO-based SVR.