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

A concept learning network based on correlation and backpropagation.

L Fu1

  • 1Dept. of Comput. & Inf. Sci., Florida Univ., Gainesville, FL.

IEEE Transactions on Systems, Man, and Cybernetics. Part B, Cybernetics : a Publication of the IEEE Systems, Man, and Cybernetics Society
|February 7, 2008
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

Management of Medullary Thyroid Carcinoma: Integrating World Health Organization Grading and Molecular Targets for Precision Therapy.

Clinical oncology (Royal College of Radiologists (Great Britain))·2026
Same author

[Non-invasive and high-precision identification of gastric precancerous lesions based on SERS and machine learning].

Zhonghua zhong liu za zhi [Chinese journal of oncology]·2026
Same author

Copper-catalyzed three-component thiocyanosulfonylation of allenes.

Chemical communications (Cambridge, England)·2026
Same author

[Recurrence of peripheral anterior synechiae following phacoemulsification combined with intraocular lens implantation and goniosynechialysis in primary angle-closure glaucoma: an intermediate and long-term analysis].

[Zhonghua yan ke za zhi] Chinese journal of ophthalmology·2026
Same author

[The impact of myocardial infarct size dynamics on left ventricular remodeling in STEMI patients after primary percutaneous coronary intervention].

Zhonghua xin xue guan bing za zhi·2025
Same author

GPER-1 Rapid Regulation Influences p-Akt Expression to Resist Stress-Induced Injuries in a Sex-Specific Manner.

Physiological research·2024
Same journal

Strategic Ability Updating in Concurrent Games by Coalitional Commitment.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2015
Same journal

Meta-Analysis of the First Facial Expression Recognition Challenge.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2012
Same journal

Adjustable model-based fusion method for multispectral and panchromatic images.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2012
Same journal

Face Feature Weighted Fusion Based on Fuzzy Membership Degree for Video Face Recognition.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2012
Same journal

A New Adaptive Fast Cellular Automaton Neighborhood Detection and Rule Identification Algorithm.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2012
Same journal

Human-arm-and-hand-dynamic model with variability analyses for a stylus-based haptic interface.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2012
See all related articles

A novel neural network integrates correlation learning with rule learning, enhancing concept learning from noisy data. This approach improves convergence when traditional methods fail, validated by theoretical and empirical analysis.

Area of Science:

  • Artificial Intelligence
  • Machine Learning
  • Neural Networks

Background:

  • Traditional neural networks often struggle with concept learning when training data is insufficient or noisy.
  • Expert systems utilize certainty factor models, but lack the adaptive learning capabilities of neural networks.

Purpose of the Study:

  • To introduce a new concept learning neural network that incorporates correlation learning.
  • To enhance the robustness of neural networks against noisy or insufficient training data.
  • To integrate the certainty factor model from expert systems into a neural network activation function.

Main Methods:

  • Developed a rule learning neural network architecture.
  • Integrated correlation learning capabilities into the network.

Related Experiment Videos

  • Utilized the certainty factor model as the network's activation function.
  • Performed theoretical analysis and empirical evaluations to validate the system.

    • Main Results:

    • The proposed network demonstrates improved convergence on concept learning tasks with noisy or insufficient data.
    • Theoretical analysis supports the network's ability to handle data imperfections.
    • Empirical evaluations confirm the effectiveness of the integrated correlation learning and certainty factor model.

    Conclusions:

    • The novel neural network architecture effectively addresses limitations in concept learning caused by data quality issues.
    • Integrating correlation learning and expert system certainty factors offers a promising direction for robust neural network design.
    • The validated system provides a more reliable approach to concept learning in challenging data scenarios.