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

Neural classifier construction using regularization, pruning and test error estimation.

Mads Hintz-Madsen1, Lars Kai Hansen, Jan Larsen

  • 1CONNECT, Department of Mathematical Modelling, Building 321, Technical University of Denmark, DK-2800, Lyngby, Denmark

Neural Networks : the Official Journal of the International Neural Network Society
|March 29, 2003
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

Longitudinal Development of Macular Ganglion Cell Layer Thickness in Children: The Copenhagen Child Cohort 2000 Eye Study.

Ophthalmic epidemiology·2026
Same author

Electron-ion equilibration in superheated gold.

Nature communications·2026
Same author

Extraction of Human Phenotype Ontology (HPO) Concepts from Clinical Notes Utilizing Large Language Models (LLM) with Model Context Protocol (MCP).

medRxiv : the preprint server for health sciences·2026
Same author

Retinal Neurophysiological Parameters in Young Patients With Type 1 Diabetes.

Investigative ophthalmology & visual science·2026
Same author

Prenatal and childhood exposure to smoking and retinal nerve fibre layer thickness: A meta-analysis of three independent birth cohorts.

Acta ophthalmologica·2026
Same author

Dim Flicker: An Endogenous Visual Percept and Its Disease Associations.

Journal of clinical medicine·2026
Same journal

Q-learning based asynchronous Boolean control networks stabilization with data loss.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

New results on prescribed-time synchronization of complex networks via intermittent control.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Variance-constrained multi-view ensemble broad network for imbalanced data.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Dynamic analysis and reliable mechanical optimization application of ring HNN effected with a memristive neuron.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

DAFF-Net: A detection and search method for small-scale low surface brightness galaxies.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Quasi-synchronization for complex networks with hybrid pinning intermittent control.

Neural networks : the official journal of the International Neural Network Society·2026
See all related articles

This study introduces a novel method for building adaptive neural classifiers using regularization. The approach employs a penalized maximum likelihood scheme and optimal brain damage pruning to select optimal network architectures for classification tasks.

Area of Science:

  • Machine Learning
  • Artificial Intelligence
  • Computational Neuroscience

Background:

  • Neural network design often involves complex architectural choices.
  • Regularization techniques are crucial for improving model generalization.
  • Estimating test error is vital for model selection and preventing overfitting.

Purpose of the Study:

  • To propose a method for constructing feed-forward neural classifiers with adaptive architectures.
  • To develop a penalized maximum likelihood scheme for deriving an entropic error measure.
  • To utilize test error estimation for optimal network architecture selection.

Main Methods:

  • A penalized maximum likelihood scheme was used to derive a modified entropic error measure.
  • An algebraic estimate of the test error was developed.

Related Experiment Videos

  • Optimal brain damage pruning was combined with the test error estimate for architecture selection.

    • Main Results:

    • The proposed method successfully constructed feed-forward neural classifiers.
    • The technique demonstrated effectiveness in selecting appropriate network architectures.
    • The scheme was validated across four distinct classification problems.

    Conclusions:

    • The developed method provides an effective approach for building regularized, adaptive neural classifiers.
    • The integration of penalized likelihood and pruning offers a robust strategy for network design.
    • The findings suggest practical applicability in various classification domains.