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 active learning for optimal generalization.

M Sugiyama1, H Ogawa

  • 1Department of Computer Science, Tokyo Institute of Technology, Meguro-ku, Tokyo, 152-8552, Japan.

Neural Computation
|December 9, 2000
PubMed
Summary

This study introduces a two-stage sampling scheme for active learning to improve model generalization by reducing bias and variance. Two novel methods, multipoint search and optimal sampling, are proposed and validated.

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
  • Statistical Learning Theory

Background:

  • Designing input signals for optimal generalization is a key challenge in machine learning, known as active learning.
  • Reducing bias and variance in models is crucial for improving predictive performance.

Purpose of the Study:

  • To propose novel active learning methods for optimal input signal design.
  • To reduce both bias and variance in machine learning models through an effective sampling scheme.

Main Methods:

  • A two-stage sampling scheme is developed to minimize bias and variance.
  • Two active learning methods are proposed: multipoint search for general models and optimal sampling for trigonometric polynomial models.

Main Results:

  • Computer simulations demonstrate the effectiveness of the multipoint search method for arbitrary models.

Related Experiment Videos

  • The optimal sampling method precisely identifies optimal sampling locations in trigonometric polynomial models.

    • Conclusions:

    • The proposed two-stage sampling scheme and active learning methods effectively enhance model generalization.
    • The developed methods offer practical solutions for designing optimal input signals in various modeling contexts.