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

Integrated feature architecture selection.

J M Steppe1, K R Bauer, S K Rogers

  • 1Dept. of Oper. Sci., Air Force Inst. of Technol., Wright-Patterson AFB, OH.

IEEE Transactions on Neural Networks
|January 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

The effect of dysmenorrhea severity and interference on reactions to experimentally-induced pain.

Frontiers in pain research (Lausanne, Switzerland)·2024
Same author

Feasibility and Accuracy of Measuring Carotid Plaque Volume (Burden) With Contrast-Enhanced Tomographic 3D Ultrasound and Ultrasound Image Fusion.

Annals of vascular surgery·2022
Same author

Optical wavelet transforms from computer-generated holography.

Applied optics·2010
Same author

Sidelobe reduction via multiaperture optical systems.

Applied optics·2010
Same author

Two-wave mixing in Ta-doped KNbO(3).

Applied optics·2010
Same author

Optical preprocessing using liquid crystal televisions.

Applied optics·2010
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 a new method for selecting optimal features and architectures for feedforward neural networks. The approach effectively reduces network complexity while maintaining prediction accuracy, demonstrated in object recognition tasks.

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Machine Learning

Background:

  • Neural networks are powerful tools for complex pattern recognition.
  • Selecting optimal features and network architecture is crucial for efficient model performance.
  • Existing methods may not adequately balance network complexity and predictive accuracy.

Purpose of the Study:

  • To develop an integrated approach for feature and architecture selection in single hidden layer-feedforward neural networks.
  • To apply a statistical model building perspective using a nonlinear regression framework.
  • To identify effective neural network models with reduced complexity and equivalent prediction accuracy.

Main Methods:

  • An integrated algorithm for feature and architecture selection was developed.

Related Experiment Videos

  • A statistical model building perspective was adopted, analyzing neural networks within a nonlinear regression framework.
  • A likelihood-ratio test statistic was employed as a model selection criterion in a sequential procedure.

    • Main Results:

    • The proposed algorithm effectively selects optimal features and architectures for neural networks.
    • The method demonstrated success in identifying reduced neural networks with equivalent prediction accuracy.
    • Application to an object recognition problem validated the algorithm's effectiveness.

    Conclusions:

    • The integrated approach provides an effective strategy for optimizing neural network design.
    • This method facilitates the development of more efficient and accurate predictive models.
    • The statistical framework offers a robust criterion for neural network model selection.