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

Side effects of normalising radial basis function networks

R Shorten1, R Murray-Smith

  • 1Daimler-Benz Research, Berlin, Germany. murray, shorten@D Bresearch-berlin.de

International Journal of Neural Systems
|May 1, 1996
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

High speed single pixel imaging using a microLED-on-CMOS light projector.

Optics express·2024
Same author

Kaon Photoproduction and the Λ Decay Parameter α_{-}.

Physical review letters·2019
Same author

The effect of multimodal and enriched feedback on SMR-BCI performance.

Clinical neurophysiology : official journal of the International Federation of Clinical Neurophysiology·2015
Same author

Clinical evaluation of BrainTree, a motor imagery hybrid BCI speller.

Journal of neural engineering·2014
Same author

Effects of the anti-androgen, bicalutamide, in a reduced life-cycle study with the fathead minnow (Pimephales promelas).

Aquatic toxicology (Amsterdam, Netherlands)·2012
Same author

Combining Brain-Computer Interfaces and Assistive Technologies: State-of-the-Art and Challenges.

Frontiers in neuroscience·2010
Same journal

Latent Space Projections and Atlases, a Cautionary Tale in Deep Neuroimaging using Autoencoders.

International journal of neural systems·2026
Same journal

Transformer-Based Anomaly Detection for Neurodegenerative Screening in MRI Images.

International journal of neural systems·2026
Same journal

Discrete Wavelet Convolution for Learnable Time-Frequency Representation with Application to Seizure Prediction.

International journal of neural systems·2026
Same journal

Automatic Seizure Detection using Hierarchical Spectral-Temporal Feature Learning with an Imbalance-Aware Transformer.

International journal of neural systems·2026
Same journal

Pyramid Vision Transformer-Enhanced Conformer Network for Epileptic Seizure Recognition Using MultiChannel EEG Signals.

International journal of neural systems·2026
Same journal

A Time-Frequency Decoupled Contrastive Learning Framework for Electroencephalography-Based Parkinson's Disease Diagnosis.

International journal of neural systems·2026
See all related articles

Normalizing Radial Basis Function (RBF) networks can cause undesirable side effects. These include altered basis function shapes, shifted maxima, and unexpected re-activation, impacting non-linear function approximation.

Area of Science:

  • Machine Learning
  • Artificial Intelligence
  • Computational Neuroscience

Background:

  • Radial Basis Function (RBF) networks are widely used for modeling applications.
  • Normalization of basis function activations ensures a partition of unity, covering the input space uniformly.
  • This normalization is a standard technique for achieving desired modeling properties.

Purpose of the Study:

  • To investigate the less desirable side effects of normalizing basis function activations in RBF networks.
  • To analyze how normalization fundamentally alters the properties of basis functions.
  • To examine the impact of these alterations on network performance and stability.

Main Methods:

  • Analysis of basis function properties under normalization.

Related Experiment Videos

  • Mathematical examination of shape changes, maxima shifts, and non-monotonic behavior.
  • Investigation of basis function re-activation phenomena.
  • Evaluation of the effects on network condition number and weights.

    • Main Results:

    • Normalization can distort basis function shapes, shifting their maxima away from centers.
    • Basis functions may no longer decrease monotonically, leading to 're-activation' far from their centers.
    • These alterations can negatively affect the stability and accuracy of non-linear function approximation.

    Conclusions:

    • The desirable partition of unity achieved by normalization comes with significant drawbacks.
    • Understanding these side effects is crucial for effective RBF network design and application.
    • Further research is needed to mitigate these issues in advanced modeling scenarios.