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

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

F L Metz1, W K Theumann

  • 1Instituto de Física, Universidade Federal do Rio Grande do Sul, Caixa Postal 15051, 91501-970 Porto Alegre, Brazil.

Physical Review. E, Statistical, Nonlinear, and Soft Matter Physics
|October 4, 2005
PubMed
Summary

This study explores pattern reconstruction versus sequence processing in a solvable neural network model. Finite noise introduces complex dynamics and new stationary states, impacting information processing.

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

Index statistical properties of sparse random graphs.

Physical review. E, Statistical, nonlinear, and soft matter physics·2015
Same author

Finite-size corrections to the spectrum of regular random graphs: An analytical solution.

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

Spectra of sparse non-hermitian random matrices: an analytical solution.

Physical review letters·2012
Same author

Spectra of sparse regular graphs with loops.

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

Area of Science:

  • Computational neuroscience
  • Statistical mechanics of neural networks

Background:

  • Investigates competition between pattern reconstruction and asymmetric sequence processing.
  • Extends prior work on parallel dynamics in neural networks far from saturation.

Purpose of the Study:

  • Analyze dynamics and stationary states in a feed-forward layered neural network model.
  • Incorporate finite stochastic noise effects from Hebbian and sequential learning rules.
  • Examine phase diagrams and quasiperiodic nonstationary solutions.

Main Methods:

  • Utilizes an exactly solvable feed-forward layered neural network model.
  • Considers binary units and patterns near saturation.
  • Introduces finite stochastic noise into the model dynamics.

Related Experiment Videos

Main Results:

  • Obtained phase diagrams detailing stationary states and quasiperiodic nonstationary solutions.
  • Identified the influence of stochastic noise on phase diagrams.
  • Analyzed the dependence of solutions on initial input overlaps.

Conclusions:

  • Finite stochastic noise significantly alters the dynamics and stationary states of the neural network model.
  • The model provides insights into the interplay between pattern reconstruction and sequence processing under noise.
  • Understanding these dynamics is crucial for designing robust artificial neural systems.