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

Attractor dynamics in feedforward neural networks.

L K Saul1, M I Jordan

  • 1AT&T Labs--Research, Florham Park, NJ 07932, USA.

Neural Computation
|August 10, 2000
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

Machine learning: Trends, perspectives, and prospects.

Science (New York, N.Y.)·2015
Same author

Computational Consequences of a Bias toward Short Connections.

Journal of cognitive neuroscience·2013
Same author

A model of the learning of arm trajectories from spatial deviations.

Journal of cognitive neuroscience·2013
Same author

Statistical modeling of biomedical corpora: mining the Caenorhabditis Genetic Center Bibliography for genes related to life span.

BMC bioinformatics·2006
Same author

Robust sparse hyperplane classifiers: application to uncertain molecular profiling data.

Journal of computational biology : a journal of computational molecular cell biology·2005
Same author

Kernel-based data fusion and its application to protein function prediction in yeast.

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing·2004
Same journal

A Model-Free Reinforcement Learning Implementation of Decision Making Under Uncertainty by Sequential Sampling.

Neural computation·2026
Same journal

DROP: Distributional and Regular Optimism and Pessimism for Reinforcement Learning.

Neural computation·2026
Same journal

Hierarchical Active Inference Using Successor Representations.

Neural computation·2026
Same journal

W-Kernel and Its Principal Space for Frequentist Evaluation of Bayesian Estimators.

Neural computation·2026
Same journal

A Hidden Markov Model-Inspired Sequence Classification Method for Hyperdimensional Computing.

Neural computation·2026
Same journal

Sparse Graphical Modeling for Electrophysiological Phase-Based Connectivity Using Circular Statistics.

Neural computation·2026
See all related articles

This study introduces attractor dynamics for probabilistic inference in feedforward neural networks, enabling unsupervised learning. The derived dynamics ensure global convergence, offering a new approach for generative models.

Area of Science:

  • Computational Neuroscience
  • Machine Learning
  • Artificial Intelligence

Background:

  • Probabilistic generative models are crucial for understanding complex data.
  • Feedforward neural networks offer a framework for parameterizing these models.
  • Probabilistic inference is essential for learning from data.

Purpose of the Study:

  • To derive and analyze attractor dynamics for probabilistic inference in feedforward neural networks.
  • To establish global convergence properties of these dynamics.
  • To demonstrate their utility for unsupervised learning signals.

Main Methods:

  • Utilizing a mean field approximation for large, layered sigmoidal networks.
  • Deriving attractor dynamics based on the mean field equations.

Related Experiment Videos

  • Employing a Lyapunov function to prove global convergence.
  • Relating unit statistics to their Markov blanket.

    • Main Results:

    • Established attractor dynamics for probabilistic inference in feedforward networks.
    • Proved global convergence of the dynamics using a Lyapunov function.
    • Demonstrated that the dynamics generate signals for unsupervised learning.
    • Provided a counterpart to existing work on symmetric networks.

    Conclusions:

    • The derived attractor dynamics offer a novel method for probabilistic inference in feedforward generative models.
    • Global convergence guarantees the stability and reliability of the learning process.
    • This approach facilitates unsupervised learning, advancing generative modeling capabilities.