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 algorithms and probability distributions in feed-forward and feed-back networks.

J J Hopfield1

  • 1Division of Chemistry, California Institute of Technology, Pasadena, CA 91125.

Proceedings of the National Academy of Sciences of the United States of America
|December 1, 1987
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

Searching for memories, Sudoku, implicit check bits, and the iterative use of not-always-correct rapid neural computation.

Neural computation·2008
Same author

Encoding for computation: recognizing brief dynamical patterns by exploiting effects of weak rhythms on action-potential timing.

Proceedings of the National Academy of Sciences of the United States of America·2004
Same author

Learning rules and network repair in spike-timing-based computation networks.

Proceedings of the National Academy of Sciences of the United States of America·2003
Same author

Simple networks for spike-timing-based computation, with application to olfactory processing.

Neuron·2003
Same author

What is a moment? Transient synchrony as a collective mechanism for spatiotemporal integration.

Proceedings of the National Academy of Sciences of the United States of America·2001
Same author

What is a moment? "Cortical" sensory integration over a brief interval.

Proceedings of the National Academy of Sciences of the United States of America·2000

This study explores learning algorithms for feed-forward and feed-back networks, particularly with noisy data. It finds that learning rules are related and statistical networks can be trained without complex Monte Carlo methods.

Area of Science:

  • Machine Learning
  • Artificial Intelligence
  • Statistical Modeling

Background:

  • Learning algorithms are crucial for deterministic and statistical networks in pattern classification and input-output relation mapping.
  • Existing methods often struggle with noisy or statistical data, limiting their application in real-world scenarios.

Purpose of the Study:

  • To examine learning algorithms for networks processing noisy or statistical data.
  • To investigate the relationship between learning rules in feed-forward and feed-back networks.
  • To explore alternative methods for training statistical networks.

Main Methods:

  • Analysis of learning algorithms applied to feed-forward deterministic and feed-back statistical networks.
  • Characterization of network behavior with noisy and statistical input data.

Related Experiment Videos

  • Investigation into the conditions under which Monte Carlo procedures can be avoided.

    • Main Results:

    • A close relationship was identified between the learning rules of feed-forward and feed-back networks in simple cases.
    • The learning problem for statistical networks can be solved without Monte Carlo procedures under specific circumstances.
    • Arbitrary learning goals for feed-forward networks can be imbued with meaningful probabilistic interpretations.

    Conclusions:

    • Learning algorithms can effectively handle noisy data by learning input-output probability distributions.
    • The study provides a more efficient approach to training statistical networks, reducing computational complexity.
    • This research bridges the gap between deterministic and statistical network learning paradigms.