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

Categorization in the symmetrically dilute Hopfield network.

P R Krebs1, W K Theumann

  • 1Instituto de Física e Matemática, Universidade Federal de Pelotas, Caixa Postal 354, 96010-900 Pelotas, RS, Brazil.

Physical Review. E, Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics
|April 24, 2002
PubMed
Summary

Gradual dilution and synaptic noise impact attractor neural network categorization. Increasing dilution enhances the stable categorization phase, improving pattern recognition in hierarchically structured data.

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

Spin-1 Hopfield model under a random field.

Physical review. E, Statistical, nonlinear, and soft matter physics·2014
Same author

Inverse melting and inverse freezing in a three-state spin-glass model with finite connectivity.

Physical review. E, Statistical, nonlinear, and soft matter physics·2013
Same author

Phase transitions in the three-state Ising spin-glass model with finite connectivity.

Physical review. E, Statistical, nonlinear, and soft matter physics·2011
Same author

Period-two cycles in a feedforward layered neural network model with symmetric sequence processing.

Physical review. E, Statistical, nonlinear, and soft matter physics·2007
Same author

Pattern reconstruction and sequence processing in feed-forward layered neural networks near saturation.

Physical review. E, Statistical, nonlinear, and soft matter physics·2005
Same author

Time evolution of the extremely diluted Blume-Emery-Griffiths neural network.

Physical review. E, Statistical, nonlinear, and soft matter physics·2004

Area of Science:

  • Computational neuroscience
  • Statistical physics

Background:

  • Attractor neural networks (ANNs) model cognitive functions like pattern recognition.
  • Hierarchically correlated patterns present challenges for ANNs due to their complex structure.

Purpose of the Study:

  • To investigate the influence of gradual dilution and synaptic noise on the categorization capabilities of a Hopfield network.
  • To understand how hierarchical pattern structures affect network performance.

Main Methods:

  • Utilized a symmetrically dilute Hopfield model with a Hebbian learning rule.
  • Employed replica-symmetric mean-field theory for an equilibrium study of the network.
  • Analyzed phase diagrams including categorization, spin-glass, and paramagnetic phases.

Related Experiment Videos

Main Results:

  • Gradual dilution significantly expands the parameter space for stable categorization.
  • Synaptic noise was incorporated into the equilibrium study.
  • Phase diagrams revealed distinct network behaviors based on dilution and noise levels.

Conclusions:

  • Gradual dilution is a key factor in enhancing the categorization ability of ANNs with hierarchical patterns.
  • The study provides insights into the robustness of ANNs against noise and dilution.
  • Findings contribute to understanding information processing in complex neural systems.