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

Noise suppression in training examples for improving generalization capability.

A Nakashima1, H Ogawa

  • 1Corporate Research & Development Center, Toshiba Corporation, Kawasaki, Japan. akiko.nakashima@toshiba.co.jp

Neural Networks : the Official Journal of the International Neural Network Society
|June 20, 2001
PubMed
Summary

Error correcting memorization learning improves generalization in supervised learning by suppressing noisy teacher signals. This method achieves projection learning levels by carefully selecting training data.

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

Pulmonary abscess caused by Cladosporium cladosporioides after receiving outpatient chemotherapy.

Journal of infection and chemotherapy : official journal of the Japan Society of Chemotherapy·2023
Same author

Identification of ABCG2 as an Exporter of Uremic Toxin Indoxyl Sulfate in Mice and as a Crucial Factor Influencing CKD Progression.

Scientific reports·2018
Same author

Temperature seasonality during fry out-migration influences the survival of hatchery-reared chum salmon Oncorhynchus keta.

Journal of fish biology·2015
Same author

Working memory training is associated with lower prefrontal cortex activation in a divergent thinking task.

Neuroscience·2013
Same author

Review: The role of autophagy in extravillous trophoblast function under hypoxia.

Placenta·2013
Same author

Inflammatory pseudotumor of the liver in association with spilled gallstones 3 years after laparoscopic cholecystectomy: report of a case.

Asian journal of endoscopic surgery·2012

Area of Science:

  • Machine Learning
  • Artificial Intelligence
  • Computer Science

Background:

  • Supervised learning often suffers from noisy teacher signals, impacting model generalization.
  • Error correcting memorization learning was developed to mitigate noise in training data.
  • Evaluating generalization capability is crucial for assessing the effectiveness of learning algorithms.

Purpose of the Study:

  • To analyze the generalization capability of error correcting memorization learning.
  • To establish conditions under which error correcting memorization learning matches projection learning.
  • To provide guidance on training set selection for optimal generalization.

Main Methods:

  • Evaluation of generalization capability using the projection learning criterion.

Related Experiment Videos

  • Derivation of a necessary and sufficient condition for equivalence between error correcting memorization learning and projection learning.
  • Analysis of the impact of noise suppression on generalization.

    • Main Results:

    • A precise condition was identified for error correcting memorization learning to achieve the same generalization as projection learning.
    • Methods for selecting training sets to meet this condition were proposed.
    • Noise suppression via error correcting memorization learning was shown to consistently enhance generalization.

    Conclusions:

    • Error correcting memorization learning offers a viable strategy for improving generalization in supervised learning.
    • The derived conditions and training set selection methods provide practical guidance for practitioners.
    • The study confirms the benefit of noise suppression for achieving high-level generalization performance.