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

Training multilayer perceptron classifiers based on a modified support vector method.

J K Suykens1, J Vandewalle

  • 1Department of Electrical Engineering, Katholieke Universiteit Leuven, ESAT-SISTA, Kardinaal Mercierlaan 94, B-3001 Leuven (Heverlee), Belgium.

IEEE Transactions on Neural Networks
|February 7, 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

Cat Scratch Colon.

Acta gastro-enterologica Belgica·2025
Same author

Short communication: Circadian variations and day-to-day variability of clinical signs used for the early diagnosis of pneumonia within and between calves.

Research in veterinary science·2023
Same author

Unusual cause of low abdominal pain.

Acta gastro-enterologica Belgica·2021
Same author

Erythema multiforme in the esophagus.

Acta gastro-enterologica Belgica·2021
Same author

The choroid plexus epithelium as a novel player in the stomach-brain axis during Helicobacter infection.

Brain, behavior, and immunity·2017
Same author

Comprehending emotional eating in obese youngsters: the role of parental rejection and emotion regulation.

International journal of obesity (2005)·2014
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 novel training method for multilayer perceptron classifiers using support vector machine (SVM) principles. The approach minimizes the Vapnik-Chervonenkis (VC) dimension for improved classification performance.

Area of Science:

  • Computational intelligence
  • Machine learning
  • Pattern recognition

Background:

  • Multilayer perceptrons (MLPs) are widely used classifiers.
  • Support Vector Machines (SVMs) offer robust classification capabilities.
  • Optimizing MLP generalization is crucial for effective pattern recognition.

Purpose of the Study:

  • To develop a new training method for single hidden layer MLPs.
  • To integrate Support Vector Machine (SVM) concepts into MLP training.
  • To enhance classifier performance by minimizing the Vapnik-Chervonenkis (VC) dimension.

Main Methods:

  • Iterative minimization of the Vapnik-Chervonenkis (VC) dimension bound.
  • Optimization of the hidden layer's interconnection matrix and bias vector.

Related Experiment Videos

  • Determination of output weights using a support vector approach, independent of Mercer's condition.

    • Main Results:

    • Successful application of the training method to a two-spiral classification problem.
    • Demonstration of an effective SVM-inspired training strategy for MLPs.
    • Validation of the VC dimension minimization technique for improved generalization.

    Conclusions:

    • The proposed method provides an effective way to train MLPs using SVM principles.
    • Minimizing the VC dimension is a viable strategy for enhancing classifier performance.
    • This approach offers an alternative to traditional MLP training and SVM formulations.