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

Learning processes in multilayer threshold nets

L Bobrowski

    Biological Cybernetics
    |November 10, 1978
    PubMed
    Summary
    This summary is machine-generated.

    This study introduces a learning algorithm for multilayer threshold nets, decomposing the problem into simpler threshold element learning. The algorithm ensures stable decision rules, even with learning perturbations.

    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

    Phenotypic features of vascular calcification in chronic kidney disease.

    Journal of internal medicine·2019
    Same author

    Induction of similarity measures and medical diagnosis support rules through separable, linear data transformations.

    Methods of information in medicine·2006
    Same author

    Diagnosis supporting rules of the Hepar system.

    Studies in health technology and informatics·2001
    Same author

    Selection of medical tests based on linear nonseparability of data sets: a study on evaluation of liver diseases tests.

    Computers in biology and medicine·1984
    Same author

    Comparison of two unsupervised algorithms.

    Biological cybernetics·1980
    Same journal

    Harmonic memory in phasor neural networks.

    Biological cybernetics·2026
    Same journal

    Correction: Decreased spinal inhibition leads to undiversified locomotor patterns.

    Biological cybernetics·2026
    Same journal

    Foundational issues of network models in biology.

    Biological cybernetics·2026
    Same journal

    Dynamical mechanisms for coordinating long-term working memory based on the precision of spike-timing in cortical neurons.

    Biological cybernetics·2026
    Same journal

    Distinct dopaminergic spike-timing-dependent plasticity rules are suited to different functional roles.

    Biological cybernetics·2026
    Same journal

    Fluctuation-response relations for a two-stage population of spiking neurons stimulated by common noise.

    Biological cybernetics·2026
    See all related articles

    Area of Science:

    • Computer Science
    • Artificial Intelligence
    • Machine Learning

    Background:

    • Multilayer threshold nets are fundamental in machine learning.
    • Learning algorithms for these nets are crucial for pattern recognition and classification tasks.
    • Existing methods may face challenges with convergence and stability, especially under noisy conditions.

    Purpose of the Study:

    • To propose a novel learning algorithm for multilayer threshold nets without feedback loops.
    • To decompose the complex net learning problem into manageable subproblems for individual threshold elements.
    • To analyze the convergence properties and stability of the proposed learning algorithm.

    Main Methods:

    • Development of a learning algorithm for threshold elements with a perceptron-like structure.

    Related Experiment Videos

  • Decomposition of the multilayer net learning task into a series of single-element learning problems.
  • Mathematical proof of decision rule stabilization for the threshold net.

    • Main Results:

    • The proposed algorithm enables learning in multilayer threshold nets with binary inputs.
    • The learning process for individual threshold elements was shown to stabilize.
    • The overall decision rule of the multilayer net was proven to stabilize in a finite number of steps.
    • Stability was demonstrated for specific classes of input distributions, including those with perturbations.

    Conclusions:

    • The developed algorithm provides an effective method for training multilayer threshold nets.
    • The decomposition strategy simplifies the learning process and guarantees convergence.
    • The algorithm's robustness to perturbations enhances its practical applicability in real-world scenarios.