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 successive overrelaxation backpropagation algorithm for neural-network training.

R De Leone1, R Capparuccia, E Merelli

  • 1Dipartimento di Matematica e Fisica, Università di Camerino, Camerino, Italy.

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

Autologous haematopoietic stem cell transplantation with an intermediate intensity conditioning regimen in multiple sclerosis: the Italian multi-centre experience.

Multiple sclerosis (Houndmills, Basingstoke, England)·2011
Same author

Modeling the cell cycle: from deterministic models to hybrid systems.

Bio Systems·2011
Same author

[Psychological issues related to multiple sclerosis].

Giornale italiano di medicina del lavoro ed ergonomia·2008
Same author

Autologous stem cell transplantation for progressive multiple sclerosis: update of the European Group for Blood and Marrow Transplantation autoimmune diseases working party database.

Multiple sclerosis (Houndmills, Basingstoke, England)·2007
Same author

Intense immunosuppression followed by autologous stem cell transplantation in severe multiple sclerosis.

Neurological sciences : official journal of the Italian Neurological Society and of the Italian Society of Clinical Neurophysiology·2006
Same author

Psychiatric disorders and depression in multiple sclerosis outpatients: impact of disability and interferon beta therapy.

Neurological sciences : official journal of the Italian Neurological Society and of the Italian Society of Clinical Neurophysiology·2005
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

A novel neural network training algorithm, a variation of backpropagation, is introduced. This method enhances convergence by utilizing updated weight values, drawing parallels with the successive overrelaxation algorithm.

Area of Science:

  • Artificial Intelligence
  • Machine Learning
  • Neural Networks

Background:

  • Classical backpropagation is a fundamental algorithm for training artificial neural networks.
  • Existing algorithms may face challenges in convergence speed and efficiency for complex network architectures.

Purpose of the Study:

  • To propose a new variation of the backpropagation algorithm for neural network training.
  • To establish the convergence properties of the proposed algorithm.

Main Methods:

  • The study introduces a modified backpropagation algorithm.
  • Convergence is established using perturbation results from Mangasarian and Solodov.
  • The algorithm's update mechanism is inspired by the successive overrelaxation (SOR) method.

Related Experiment Videos

Main Results:

  • The proposed algorithm demonstrates convergence, validated through theoretical analysis.
  • The method's similarity to SOR suggests potential efficiency gains in weight updates.
  • Utilizing the most recently computed weight values optimizes the training process.

Conclusions:

  • The novel backpropagation variation offers a theoretically sound approach to neural network training.
  • The algorithm's convergence properties are rigorously established.
  • The SOR-inspired update strategy provides a promising direction for enhancing neural network training efficiency.