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

Diffusion approximation of frequency sensitive competitive learning.

A S Galanopoulos1, R L Moses, S C Ahalt

  • 1Dept. of Electr. Eng., Ohio State Univ., Columbus, OH.

IEEE Transactions on Neural Networks
|January 1, 1997
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

Generation of 10 patient-specific induced pluripotent stem cells (iPSCs) to model Pitt-Hopkins Syndrome.

Stem cell research·2020
Same author

Optical implementation and applications of closest-vector selection in neural networks.

Applied optics·2010
Same author

Extracellular matrix molecules improve periodontal ligament cell adhesion to anorganic bone matrix.

Journal of dental research·2001
Same author

Evaluation of metallic and polymeric biomaterial surface energy and surface roughness characteristics for directed cell adhesion.

Tissue engineering·2001
Same author

Unusual otolaryngic presentations of ameloblastoma.

Otolaryngology--head and neck surgery : official journal of American Academy of Otolaryngology-Head and Neck Surgery·1999
Same author

Endoscopic transseptal transsphenoidal hypophysectomy with three-dimensional intraoperative localization technology.

The Laryngoscope·1999
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 analyzes the frequency sensitive competitive learning (FSCL) algorithm

Area of Science:

  • Computational neuroscience
  • Machine learning algorithms

Background:

  • Competitive learning algorithms are crucial for unsupervised learning.
  • Understanding convergence properties is essential for algorithm reliability.

Purpose of the Study:

  • To analyze the convergence of the frequency sensitive competitive learning (FSCL) algorithm.
  • To establish conditions for the convergence of FSCL to a local equilibrium.
  • To broaden the understanding of convergence characteristics in learning algorithms.

Main Methods:

  • Approximation of the final FSCL learning phase using a diffusion process.
  • Mathematical analysis employing the Fokker-Plank equation.
  • Comparison with convergence analysis of Kohonen's self-organizing map.

Related Experiment Videos

Main Results:

  • Sufficient and necessary conditions for convergence to a local equilibrium were identified.
  • Convergence conditions were found to depend solely on the learning rate.
  • These conditions were shown to be equivalent to previously described weak convergence criteria.

Conclusions:

  • The study provides a rigorous convergence analysis for the FSCL algorithm.
  • The findings extend the applicability of established convergence characteristics to a broader class of algorithms.
  • The learning rate is identified as the sole determinant of convergence in the analyzed FSCL phase.