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

Variational learning in nonlinear gaussian belief networks

B J Frey1, G E Hinton

  • 1Beckman Institute for Advanced Science & Technology, University of Illinois at Urbana-Champaign, 405 North Mathew Avenue, Urbana IL 61801, USA. bfrey@dendrite.beckman.uiuc.edu

Neural Computation
|February 9, 1999
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

Glove-TalkII--a neural-network interface which maps gestures to parallel formant speech synthesizer controls.

IEEE transactions on neural networks·2008
Same author

Comparing sequence and expression for predicting microRNA targets using GenMiR3.

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing·2008
Same author

Reducing the dimensionality of data with neural networks.

Science (New York, N.Y.)·2006
Same author

Computation by neural networks.

Nature neuroscience·2000
Same author

SMEM algorithm for mixture models.

Neural computation·2000
Same author

Variational learning for switching state-space models.

Neural computation·2000
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 a novel framework for stochastic neural networks to handle nonlinear perceptual tasks. The approach uses a variational method for improved feature extraction and pattern classification, outperforming existing methods in digit recognition.

Area of Science:

  • Artificial Intelligence
  • Machine Learning
  • Computer Vision

Background:

  • Perceptual tasks like vision and speech recognition are framed as inference problems.
  • Independent Component Analysis highlights the need for inferring continuous latent variables.
  • Existing methods often rely on linear relationships, insufficient for complex nonlinear perception.

Purpose of the Study:

  • To present a unifying framework for stochastic neural networks with nonlinear latent variables.
  • To develop a general variational method for maximizing the likelihood of training data.
  • To apply the framework to visual feature extraction and pattern classification.

Main Methods:

  • Developed stochastic neural networks with nonlinear latent variables by applying nonlinearities to linear Gaussian units.

Related Experiment Videos

  • Implemented a general variational method to optimize a lower bound on the data likelihood.
  • Evaluated performance on visual feature extraction and handwritten digit recognition tasks.

    • Main Results:

    • Demonstrated the framework's effectiveness on two visual feature extraction problems.
    • Showcased the variational method's utility for pattern classification.
    • Achieved competitive performance in handwritten digit recognition compared to other methods.

    Conclusions:

    • The proposed framework offers a unified approach for nonlinear perceptual inference.
    • The variational method provides a robust technique for training and applying these networks.
    • Nonlinear stochastic neural networks show significant potential for complex pattern recognition tasks.