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

Replicator neural networks for universal optimal source coding.

R Hecht-Nielsen

    Science (New York, N.Y.)
    |September 29, 1995
    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

    Counterpropagation networks.

    Applied optics·2010
    Same author

    Nearest matched filter classification of spatiotemporal patterns.

    Applied optics·2010
    Same author

    Conditional probability density function estimation with sigmoidal neural networks.

    IEEE transactions on neural networks·2008
    Same author

    A biologically motivated solution to the cocktail party problem.

    Neural computation·2001
    Same author

    A supervised learning neural network coprocessor for soft-decision maximum-likelihood decoding.

    IEEE transactions on neural networks·1995
    Same journal

    A native sulfur deposit in Gale crater, Mars.

    Science (New York, N.Y.)·2026
    Same journal

    Coordinated demise of harmful algal blooms.

    Science (New York, N.Y.)·2026
    Same journal

    Genetic effects put into context.

    Science (New York, N.Y.)·2026
    Same journal

    Bacteria share proteins to survive antibiotics.

    Science (New York, N.Y.)·2026
    Same journal

    Impacts shaped Earth's first continents.

    Science (New York, N.Y.)·2026
    Same journal

    Erratum for the Report "Covalently bonded single-molecule junctions with stable and reversible photoswitched conductivity" by C. Jia <i>et al</i>.

    Science (New York, N.Y.)·2026
    See all related articles

    Replicator neural networks achieve optimal data compression by minimizing reconstruction error. These networks also discover unique coordinate systems for data manifolds, enhancing data representation.

    Area of Science:

    • Artificial Intelligence
    • Machine Learning
    • Information Theory

    Background:

    • Replicator neural networks self-organize by treating inputs as desired outputs, creating internal compressed data representations.
    • Previous research indicated potential for optimal data compression in these networks.

    Purpose of the Study:

    • To demonstrate that replicator neural networks can achieve optimal data compression for arbitrary data sources.
    • To introduce a new data source model, data manifolds, and analyze network behavior within this framework.

    Main Methods:

    • Utilizing a theorem proving optimal data compression via minimization of mean squared reconstruction error.
    • Training replicator networks on raw data examples.
    • Introducing and analyzing the properties of data manifolds.

    Related Experiment Videos

    Main Results:

    • A theorem confirms that minimizing mean squared reconstruction error enables optimal data compression for diverse data sources.
    • The introduction of data manifolds as a generalized data source model.
    • A second theorem shows that optimal-compression networks uniquely define natural coordinate systems for data manifolds in a key limiting case.

    Conclusions:

    • Replicator neural networks are theoretically capable of optimal data compression.
    • The concept of data manifolds provides a novel framework for understanding data sources.
    • These networks naturally uncover underlying structures within data, forming unique coordinate systems.