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

Neuro-fuzzy feature evaluation with theoretical analysis.

R K. De1, J Basak, S K. Pal

  • 1Machine Intelligence Unit, Indian Statistical Institute, Calcutta, India

Neural Networks : the Official Journal of the International Neural Network Society
|March 29, 2003
PubMed
Summary

This study introduces a novel fuzzy set theoretic feature evaluation index and a neuro-fuzzy algorithm to determine feature importance for improved classification accuracy. The method effectively assesses features individually and in groups across diverse datasets.

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 role of microenvironmental pH regulation on colorectal cancer cells - a preliminary study.

Journal of physiology and pharmacology : an official journal of the Polish Physiological Society·2025
Same author

Deciphering miRNA-lncRNA-mRNA interaction through experimental validation of miRNAs, lncRNAs, and miRNA targets on mRNAs in Cajanus cajan.

Plant biology (Stuttgart, Germany)·2024
Same author

The dynamics of violence in everyday living - an Indian context.

The International journal of psycho-analysis·2023
Same author

Imatinib mesylate therapy in patients of chronic myeloid leukemia with Philadelphia chromosome positive: an experience from eastern India.

Indian journal of hematology & blood transfusion : an official journal of Indian Society of Hematology and Blood Transfusion·2013
Same author

Stem cell transplant: An experience from eastern India.

Indian journal of medical and paediatric oncology : official journal of Indian Society of Medical & Paediatric Oncology·2013
Same author

Plant-pathogen interactions: what microarray tells about it?

Molecular biotechnology·2011

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Machine Learning

Background:

  • Feature evaluation is crucial for effective machine learning models.
  • Traditional methods may not adequately capture complex feature relationships.
  • Fuzzy set theory offers a framework for handling uncertainty in feature data.

Purpose of the Study:

  • To develop a fuzzy set theoretic feature evaluation index.
  • To introduce a connectionist model for evaluating the index.
  • To propose a neuro-fuzzy algorithm for optimizing feature weights.

Main Methods:

  • Development of a weighted membership function for improved class modeling.
  • Implementation of a neuro-fuzzy algorithm to determine optimal weighting coefficients.

Related Experiment Videos

  • Theoretical analysis of the evaluation index's properties (bounds, monotonicity).

    • Main Results:

    • The feature evaluation index demonstrates theoretical bounds and monotonic increase with lower bounds.
    • A clear relationship is established between the evaluation index, interclass distance, and weighting coefficients.
    • The algorithm's effectiveness is validated on speech, Iris, medical, and mango-leaf datasets.

    Conclusions:

    • The proposed neuro-fuzzy approach provides an effective method for feature evaluation.
    • The method accurately assesses feature importance, considering both individual and group dependencies.
    • Results are corroborated by scatter diagrams and k-Nearest Neighbors (k-NN) classification.