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

Universal approximation using incremental constructive feedforward networks with random hidden nodes.

Guang-Bin Huang, Lei Chen, Chee-Kheong Siew

    IEEE Transactions on Neural Networks
    |July 22, 2006
    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

    Artificial Intelligence Diagnosis of Obstructive Sleep Apnea Using Overnight Pulse Oximetry: A Systematic Review and Bayesian Meta-Analysis.

    Journal of medical Internet research·2026
    Same author

    Artificial Intelligence Diagnosis of Obstructive Sleep Apnoea using Overnight Pulse Oximetry: A Systematic Review and Bayesian Meta-Analysis.

    Journal of medical Internet research·2026
    Same author

    Awake speech recordings for machine learning diagnosis of obstructive sleep apnea: a Bayesian meta-analysis.

    Journal of clinical sleep medicine : JCSM : official publication of the American Academy of Sleep Medicine·2026
    Same author

    Artificial intelligence without restriction surpassing human intelligence with probability one: Theoretical insight into secrets of the brain with AI twins of the brain.

    Neurocomputing·2025
    Same author

    N7-methylguanosine-related gene decapping scavenger enzymes as a novel biomarker regulating epithelial cell function in diabetic foot ulcers.

    World journal of diabetes·2025
    Same author

    Multichannel machine learning for polysomnographic diagnosis of obstructive sleep apnea: a Bayesian meta-analysis.

    European archives of oto-rhino-laryngology : official journal of the European Federation of Oto-Rhino-Laryngological Societies (EUFOS) : affiliated with the German Society for Oto-Rhino-Laryngology - Head and Neck Surgery·2025

    This study demonstrates that single-hidden-layer feedforward networks (SLFNs) can achieve universal approximation by randomly selecting hidden nodes and only adjusting output weights. This simplifies training, even for networks with nondifferentiable activation functions.

    Area of Science:

    • Artificial Intelligence
    • Machine Learning
    • Neural Networks

    Background:

    • Conventional neural network theories posit that single-hidden-layer feedforward networks (SLFNs) require all parameters to be adjustable for universal approximation.
    • Tuning all parameters in SLFNs often leads to complex and inefficient learning, especially with nondifferentiable activation functions.

    Purpose of the Study:

    • To prove that SLFNs can be universal approximators by randomly selecting hidden nodes and only adjusting output weights.
    • To develop an efficient and automatic training method for SLFNs.

    Main Methods:

    • An incremental constructive method is employed to demonstrate the universal approximation capability of SLFNs.
    • The study focuses on randomly initializing hidden nodes and exclusively training the output weights.

    Related Experiment Videos

    Main Results:

    • The proposed method allows SLFNs with various activation functions (continuous, nondifferentiable, piecewise continuous) to act as universal approximators.
    • This approach simplifies the training process, making it efficient and automatic, eliminating the need for manual parameter tuning.

    Conclusions:

    • SLFNs can achieve universal approximation with a simplified training approach, randomly initializing hidden nodes and training only output weights.
    • The incremental constructive method offers an efficient, automatic solution for training SLFNs, applicable to diverse activation functions.