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

Continuous-valued probabilistic behavior in a VLSI generative model.

Hsin Chen1, Patrice C D Fleury, Alan F Murray

  • 1Department of Electrical Engineering, National Tsing-Hua University, Hsin-Chu 30055, Taiwan. hchen@ee.nthu.edu.tw

IEEE Transactions on Neural Networks
|May 26, 2006
PubMed
Summary

This study introduces a hardware-friendly VLSI implementation of the continuous restricted Boltzmann machine (CRBM) for modeling continuous data. The system integrates on-chip training, enabling exploration of probabilistic computation in VLSI.

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

Distinct metabolic dependency on BCAA defines cancer stemness and malignancy in human triple-negative breast cancer.

Cell reports·2026
Same author

Scanning Bessel beam microscopy with a protected and corrective objective for solvent-cleared large samples.

iScience·2026
Same author

Suppression of visceral nociception by C-fiber transmission block using temporal interference sinusoidal stimulation.

Pain reports·2026
Same author

Risks of cardiovascular disease and cerebrovascular disease following kidney transplantation: A nationwide, population-based cohort study.

International journal of medical sciences·2025
Same author

Nipocalimab, an immunoselective FcRn blocker that lowers IgG and has unique molecular properties.

mAbs·2025
Same author

Rapid lightsheet fluorescence imaging of whole Drosophila brains at nanoscale resolution by potassium acrylate-based expansion microscopy.

Nature communications·2024

Area of Science:

  • Computer Engineering
  • Artificial Intelligence
  • Machine Learning

Background:

  • Probabilistic generative models are crucial for understanding complex data distributions.
  • Continuous restricted Boltzmann machines (CRBMs) offer a powerful approach for modeling continuous-valued data.
  • Efficient hardware implementations are needed to realize the potential of CRBMs in real-world applications.

Purpose of the Study:

  • To present the Very Large-Scale Integration (VLSI) implementation of a continuous restricted Boltzmann machine (CRBM).
  • To demonstrate a hardware-amenable training algorithm for CRBMs.
  • To explore computation with continuous-valued probabilistic behavior in VLSI.

Main Methods:

  • Designed and implemented a CRBM system using VLSI technology.

Related Experiment Videos

  • Incorporated stochastic neurons with noise-mediated continuous-valued probabilistic behavior.
  • Integrated on-chip training circuits for autonomous learning.

    • Main Results:

    • Successfully modeled continuous-valued data distributions using the VLSI CRBM.
    • Demonstrated the ability to regenerate continuous-valued data distributions.
    • Identified and discussed performance limitations of the implemented system.

    Conclusions:

    • The VLSI CRBM provides a viable platform for hardware-based probabilistic modeling of continuous data.
    • On-chip training circuits facilitate practical applications of CRBMs in VLSI.
    • Further research is needed to overcome performance limitations and enhance CRBM capabilities.