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

Parallel sequential minimal optimization for the training of support vector machines.

L J Cao1, S S Keerthi, Chong-Jin Ong

  • 1Financial Studies, Fudan University, ShangHai, PR China. ljcao@fudan.edu.cn

IEEE Transactions on Neural Networks
|July 22, 2006
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

Precise Measurement of the Chromoelectric Dipole Moment of the Charm Quark.

Physical review letters·2026
Same author

[An interview with Dr. Guo Xiumei on Sino-Japanese academic exchange].

Zhonghua yi shi za zhi (Beijing, China : 1980)·2026
Same author

Precise Measurement of Matter-Antimatter Asymmetry with Entangled Hyperon-Antihyperon Pairs.

Physical review letters·2026
Same author

Observation of Λ[over ¯]p→K^{+}π^{+}π^{-}π^{0} and Λ[over ¯]p→K^{+}π^{+}π^{-}2π^{0}.

Physical review letters·2026
Same author

First Measurement of the D_{s}^{+}→K^{0}μ^{+}ν_{μ} Decay.

Physical review letters·2026
Same author

Observation of the Electromagnetic Radiative Decays of the Λ(1520) and Λ(1690) to γΣ^{0}.

Physical review letters·2026
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 introduces a parallel implementation of Sequential Minimal Optimization (SMO) for Support Vector Machine (SVM) training. The parallel approach significantly speeds up computation for large datasets using Message Passing Interface (MPI).

Area of Science:

  • Computer Science
  • Machine Learning
  • Artificial Intelligence

Background:

  • Sequential Minimal Optimization (SMO) is a widely used algorithm for training Support Vector Machines (SVMs).
  • Training SVMs on large datasets with SMO can be computationally intensive and time-consuming.
  • Existing SMO algorithms face challenges in efficiently handling large-scale machine learning problems.

Purpose of the Study:

  • To develop and evaluate a parallel implementation of the SMO algorithm for training SVMs.
  • To address the computational bottlenecks associated with training SVMs on large datasets.
  • To improve the efficiency and scalability of SVM training using parallel processing.

Main Methods:

  • The proposed parallel SMO algorithm is implemented using the Message Passing Interface (MPI).

Related Experiment Videos

  • The training dataset is partitioned into smaller subsets.
  • Multiple CPU processors are utilized to process these subsets concurrently.

    • Main Results:

    • Significant speedups were observed when training on the adult and MNIST datasets with multiple processors.
    • The parallel SMO demonstrated satisfactory performance on the Web dataset.
    • The implementation effectively leverages parallel processing to reduce SVM training time.

    Conclusions:

    • The parallel implementation of SMO using MPI offers a viable solution for accelerating SVM training on large datasets.
    • This approach enhances the scalability of SVM training, making it more practical for real-world applications.
    • The findings suggest that parallelization is a key strategy for overcoming computational limitations in large-scale machine learning.