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

Automatic basis selection techniques for RBF networks.

Ali Ghodsi1, Dale Schuurmans

  • 1School of Computer Science, University of Waterloo, 200 University Avenue West, Waterloo, Ont., Canada N2L 3G1. aghodsib@cs.uwaterlo.ca

Neural Networks : the Official Journal of the International Neural Network Society
|July 10, 2003
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

Learning Chemotherapy Drug Action via Universal Physics-Informed Neural Networks.

Pharmaceutical research·2025
Same author

GraphPI: Efficient Protein Inference with Graph Neural Networks.

Journal of proteome research·2024
Same author

Influence of photothermal nanomaterials localization within the electrospun membrane structure on purification of saline oily wastewater based on photothermal vacuum membrane distillation.

Journal of environmental management·2024
Same author

Fine-Tuning and training of densenet for histopathology image representation using TCGA diagnostic slides.

Medical image analysis·2021
Same author

Energy and exergy analysis of a novel turbo-compounding system for supercharging and mild hybridization of a gasoline engine.

Journal of thermal analysis and calorimetry·2020
Same author

Deep learning enables de novo peptide sequencing from data-independent-acquisition mass spectrometry.

Nature methods·2018
Same journal

Exploiting audio-visual modalities in videos: Object detection via multi-stage bilateral coupling network.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Reliability-aware modality completion with cross-modal distillation for federated learning with missing modalities.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

IGFD-Net: Illumination-guided frequency decoupling for polarization image fusion.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Multiple-Strategies dung beetle optimizer and its applications in engineering optimization and bankruptcy prediction.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Aggregating global-scale pixel-wise forgery cues within a graph.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Finite-Time intermittent control for secure synchronization of Neutral-Type stochastic delayed neural networks under aperiodic DoS attacks.

Neural networks : the official journal of the International Neural Network Society·2026
See all related articles

This study introduces a new criterion to determine the optimal number of basis functions for radial basis function (RBF) neural networks. This method balances bias and variance for improved generalization performance in RBF networks.

Area of Science:

  • Machine Learning
  • Artificial Intelligence
  • Computational Neuroscience

Background:

  • Radial basis function (RBF) neural networks require careful selection of basis functions for optimal performance.
  • Generalization performance, crucial for prediction on new data, is influenced by the trade-off between bias and variance.
  • Overly simple or overly complex RBF networks lead to poor generalization.

Purpose of the Study:

  • To propose a generic criterion for determining the optimum number of basis functions in RBF neural networks.
  • To provide a method for achieving the best compromise between bias and variance.
  • To validate the proposed criterion experimentally.

Main Methods:

  • Derivation of an analytical criterion using Stein's unbiased risk estimator.

Related Experiment Videos

  • Consideration of two noise scenarios: known and unknown noise.
  • Experimental validation of the criterion's efficacy.

    • Main Results:

    • The proposed criterion effectively determines the optimal number of basis functions for RBF networks.
    • The method demonstrates efficacy in both known and unknown noise conditions.
    • Empirical comparison shows competitive or superior performance against cross-validation and Bayesian Information Criterion (BIC).

    Conclusions:

    • The developed criterion offers a robust approach to optimizing RBF network architecture.
    • This method enhances the generalization capability of RBF neural networks.
    • The criterion provides a valuable tool for model selection in RBF network design.