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 noisy self-organizing neural network with bifurcation dynamics for combinatorial optimization.

Terence Kwok1, Kate A Smith

  • 1School of Business Systems, Faculty of Information Technology, Monash University, Clayton, Victoria 3168, Australia. terence.kwok@infotech.monash.edu.au

IEEE Transactions on Neural Networks
|September 25, 2004
PubMed
Summary

This study enhances self-organizing neural networks (SONNs) for combinatorial optimization problems (COPs) by introducing weight normalization with bifurcation dynamics and additive noise. These improvements boost convergence and solution quality for complex optimization tasks.

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

The Frequency of Signs of Meibomian Gland Dysfunction in Children with Epidermolysis Bullosa.

Ophthalmology·2016
Same author

Optimization via intermittency with a self-organizing neural network.

Neural computation·2005
See all related articles

Area of Science:

  • Computational neuroscience
  • Artificial intelligence
  • Optimization theory

Background:

  • Existing self-organizing neural networks (SONNs) face limitations in convergence speed and solution quality for general "0-1" combinatorial optimization problems (COPs).
  • Addressing these limitations requires novel network architectures and dynamics.

Purpose of the Study:

  • To improve the performance of SONNs for solving "0-1" COPs.
  • To investigate the impact of weight normalization with bifurcation dynamics and additive neuronal noise on optimization.
  • To provide theoretical and experimental insights into SONN behavior and parameter dependencies.

Main Methods:

  • Development and analysis of a SONN incorporating an efficient weight normalization process with bifurcation dynamics.

Related Experiment Videos

  • Inclusion of additive noise in network neurons.
  • Experimental validation using the N-queen problem to study parameter effects, including annealing schedules.
  • Derivation of an equilibrium model for the SONN with neuronal weight normalization.
  • Analysis of dynamical systems properties, including period-doubling bifurcations and strange attractors.

    • Main Results:

    • The enhanced SONN demonstrates improved convergence and solution quality compared to existing methods.
    • Weight normalization with bifurcation dynamics explains observed high feasibility regions in parameter space.
    • The addition of random noise effectively reduces oscillations, leading to better solution feasibility.
    • Dynamical systems analysis reveals period-doubling bifurcations to chaos under specific conditions, with annealing temperature as a key parameter.

    Conclusions:

    • The proposed SONN architecture with weight normalization and additive noise offers a significant advancement for solving "0-1" COPs.
    • Bifurcation dynamics and noise are crucial for understanding and optimizing SONN performance.
    • The study provides a theoretical framework and experimental evidence for the effectiveness of the enhanced SONN model.