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

Incremental projection learning for optimal generalization.

M Sugiyama1, H Ogawa

  • 1Department of Computer Science, Graduate School of Information Science and Engineering, Tokyo Institute of Technology, Japan. sugi@og.cs.titech.ac.jp

Neural Networks : the Official Journal of the International Neural Network Society
|February 24, 2001
PubMed
Summary

This study introduces incremental projection learning, a novel method that enhances neural network generalization. It achieves results comparable to batch learning, overcoming limitations of existing incremental learning techniques.

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

Atopic dermatitis: studies of skin permeability and effectiveness of topical PUVA treatment.

Pediatric dermatology·1992
Same author

Stimulation of keratinocyte migration by growth factors.

The Journal of dermatology·1992
Same author

A wound healing model using healing-impaired diabetic mice.

The Journal of dermatology·1992
Same author

The response to drug therapy in unstable angina on the basis of coronary angiography findings.

Japanese circulation journal·1992
Same author

[The pathogenic role of Chlamydia trachomatis in otitis media with effusion].

Nihon Jibiinkoka Gakkai kaiho·1992
Same author

[Prospective randomized trials comparing hyperfractionated radiotherapy with conventional radiotherapy in stage III non-small cell lung cancer].

Nihon Igaku Hoshasen Gakkai zasshi. Nippon acta radiologica·1992

Area of Science:

  • Machine Learning
  • Artificial Intelligence
  • Neural Networks

Background:

  • Improving neural network generalization after initial training is crucial.
  • Current incremental learning methods often yield suboptimal generalization compared to batch learning.
  • Adding training data is a common but sometimes insufficient approach.

Purpose of the Study:

  • To present a novel incremental projection learning method.
  • To address the poor generalization capability of existing incremental learning techniques.
  • To achieve generalization performance equivalent to batch learning methods.

Main Methods:

  • Developed an incremental projection learning algorithm.
  • Incorporated noise handling within the incremental learning framework.

Related Experiment Videos

  • Validated the method through computer simulations.

    • Main Results:

    • The proposed method achieves identical learning results to batch projection learning.
    • Demonstrated effective generalization capability in the presence of noise.
    • Computer simulations confirmed the method's practical effectiveness.

    Conclusions:

    • Incremental projection learning offers a viable solution for enhancing neural network generalization.
    • The method successfully bridges the performance gap between incremental and batch learning.
    • This approach provides a robust way to update neural networks with new data without compromising performance.