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

Orthogonal least squares learning algorithm for radial basis function networks.

S Chen1, C N Cowan, P M Grant

  • 1Dept. of Electr. Eng., Edinburgh Univ.

IEEE Transactions on Neural Networks
|January 1, 1991
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

Fractionation of copper activation products in debris samples from the National Ignition Facility.

Applied radiation and isotopes : including data, instrumentation and methods for use in agriculture, industry and medicine·2018
Same author

Distribution of collected target debris using the large area solid debris radiochemistry collector.

The Review of scientific instruments·2018
Same author

Rapid improvement in beliefs, mood, and performance following an experimental success experience in an analogue test of recovery-oriented cognitive therapy.

Psychological medicine·2017
Same author

Recent advances and results from the solid radiochemistry nuclear diagnostic at the National Ignition Facility.

The Review of scientific instruments·2016
Same author

From neurocognition to community participation in serious mental illness: the intermediary role of dysfunctional attitudes and motivation.

Psychological medicine·2016
Same author

Note: Radiochemical measurement of fuel and ablator areal densities in cryogenic implosions at the National Ignition Facility.

The Review of scientific instruments·2015
Same journal

Universal perceptron and DNA-like learning algorithm for binary neural networks: LSBF and PBF implementations.

IEEE transactions on neural networks·2013
Same journal

Guest editorial: special section on white box nonlinear prediction models.

IEEE transactions on neural networks·2011
Same journal

Data-based fault-tolerant control of high-speed trains with traction/braking notch nonlinearities and actuator failures.

IEEE transactions on neural networks·2011
Same journal

Guest editorial: special section on data-based control, modeling, and optimization.

IEEE transactions on neural networks·2011
Same journal

Neural network-based multiple robot simultaneous localization and mapping.

IEEE transactions on neural networks·2011
Same journal

Data-driven model-free adaptive control for a class of MIMO nonlinear discrete-time systems.

IEEE transactions on neural networks·2011
See all related articles

This study introduces an improved learning method for radial basis function networks using orthogonal least-squares. This approach rationally selects centers to enhance signal processing performance and avoid common drawbacks.

Area of Science:

  • Machine Learning
  • Signal Processing
  • Artificial Neural Networks

Background:

  • Radial basis function (RBF) networks are effective alternatives to traditional two-layer neural networks in signal processing.
  • Current RBF network learning algorithms often rely on random center selection and singular-value decomposition, which have significant drawbacks.
  • Arbitrary selection of centers in RBF networks leads to suboptimal performance and potential numerical instability.

Purpose of the Study:

  • To propose a novel, rational learning procedure for RBF networks.
  • To address the limitations of arbitrary center selection in existing RBF network training algorithms.
  • To develop an efficient and numerically stable method for fitting RBF networks.

Main Methods:

  • The study proposes an alternative learning procedure based on the orthogonal least-squares (OLS) method.

Related Experiment Videos

  • OLS selects RBF centers iteratively, maximizing the explained variance of the output at each step.
  • This method avoids numerical ill-conditioning problems associated with traditional approaches.

    • Main Results:

    • The OLS-based procedure constructs RBF networks by selecting centers one by one in a rational manner.
    • Each selected center contributes maximally to the explained variance or energy of the desired output.
    • The proposed method demonstrates efficiency and simplicity in fitting RBF networks, as shown in signal processing examples.

    Conclusions:

    • The orthogonal least-squares learning strategy offers a superior method for training radial basis function networks.
    • This approach provides a rational and efficient alternative to existing learning algorithms, particularly in signal processing applications.
    • The method ensures numerical stability and optimal network construction through a systematic center selection process.