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

A dynamical system perspective of structural learning with forgetting.

D A Miller1, J M Zurada

  • 1Department of Electrical and Computer Engineering, Western Michigan University, Kalamazoo, MI 49008, USA.

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

Large-scale single-molecule analysis of tau proteoforms.

bioRxiv : the preprint server for biology·2025
Same author

Cadmium phytoextraction by Helianthus annuus (sunflower), Brassica napus cv Wichita (rapeseed), and Chyrsopogon zizanioides (vetiver).

Chemosphere·2020
Same author

A Multicenter Pilot Study on the Clinical Utility of Computational Modeling for Flow-Diverter Treatment Planning.

AJNR. American journal of neuroradiology·2019
Same author

Demographic and mortality analysis of hospitalized children at a referral hospital in Addis Ababa, Ethiopia.

BMC pediatrics·2016
Same author

Prenatal diagnosis of intraventricular hemorrhage in a fetus with congenital diaphragmatic hernia using magnetic resonance imaging.

Minerva ginecologica·2015
Same author

Psychiatry clerkship sites : factors that make a difference in student performance.

Academic psychiatry : the journal of the American Association of Directors of Psychiatric Residency Training and the Association for Academic Psychiatry·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 continuous dynamical system for structural learning with forgetting, demonstrating Laplace regularization

Area of Science:

  • Artificial Intelligence
  • Machine Learning
  • Computational Neuroscience

Background:

  • Structural learning with forgetting utilizes Laplace regularization for skeletal artificial neural networks.
  • Existing methods focus on static regularization parameters.

Purpose of the Study:

  • To develop a continuous dynamical system model for regularization.
  • To generalize the regularization parameter as a time-varying function.
  • To compare Laplace and Gaussian regularization within this dynamic framework.

Main Methods:

  • Developed a continuous dynamical system model for regularization.
  • Analyzed a Laplace regularizer with a quadratic error surface.
  • Solved linear systems in different weight space regions.

Related Experiment Videos

  • Compared Laplace and Gaussian regularization effects.

    • Main Results:

    • Laplace regularization acts as a control input, while Gaussian regularization modifies eigenvalues.
    • Both regularizers impact less important weight space directions more significantly.
    • The dynamic model provides analytic results for Laplace regularization.

    Conclusions:

    • The developed dynamic system offers new insights into regularization.
    • Laplace regularization demonstrates superiority over Gaussian regularization in this context.
    • Time-varying regularization parameters offer a more nuanced approach to structural learning.