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

On computing the fuzzifier in downward arrow FLVQ: a data driven approach.

L B Romdhane1, B Ayeb, S Wang

  • 1Department of Mathematics and Computer Sciences, Faculty of Sciences, University of Sherbrooke, Qc, Canada. benromdh@dmi.usherb.ca

International Journal of Neural Systems
|May 30, 2002
PubMed
Summary

This study introduces a new heuristic method for setting the fuzzifier parameter in fuzzy clustering algorithms. This approach improves cluster analysis by adapting the fuzzifier to the specific dataset, enhancing performance in applications like image compression.

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

A novel G protein-coupled P2 purinoceptor (P2Y3) activated preferentially by nucleoside diphosphates.

Molecular pharmacology·1996
Same author

Endometrial lymphomyeloid cells in abnormal uterine bleeding due to levonorgestrel (Norplant).

Human reproduction (Oxford, England)·1996
Same author

Hydrophobicity and sorption of chlorophenolates to lipid membranes.

Chemosphere·1996
Same author

Inhibition of lipid peroxidation in low-density lipoprotein by the flavonoid myrigalone B and ascorbic acid.

Biochemical pharmacology·1996
Same author

The fourth EF-hand of calmodulin and its helix-loop-helix components: impact on calcium binding and enzyme activation.

Biochemistry·1996
Same author

Antiretroviral agents as inhibitors of both human immunodeficiency virus type 1 integrase and protease.

Journal of medicinal chemistry·1996

Area of Science:

  • Computer Science
  • Data Science
  • Artificial Intelligence

Background:

  • Fuzzy clustering offers advantages over traditional methods, particularly with overlapping data.
  • Existing fuzzy c-means algorithms rely on a critical parameter, the fuzzifier, with no established optimal setting.
  • The fuzzifier exponent significantly impacts cluster determination in fuzzy clustering.

Purpose of the Study:

  • To develop a novel heuristic scheme for determining the optimal fuzzifier parameter in fuzzy clustering.
  • To create a method that dynamically links the fuzzifier to the dataset being analyzed.
  • To address the lack of theoretical guidance for setting the fuzzifier in fuzzy clustering algorithms.

Main Methods:

  • Development of a heuristic scheme for fuzzifier determination.

Related Experiment Videos

  • Integration of data-set-specific interactions into the fuzzifier selection process.
  • Application of the proposed method to clustering benchmark datasets (IRIS) and image compression tasks.

    • Main Results:

    • The proposed heuristic scheme demonstrated effective determination of the fuzzifier parameter.
    • Experimental results showed good performance in clustering the IRIS dataset.
    • The method proved successful in codebook design for image compression applications.

    Conclusions:

    • The developed heuristic scheme provides a practical approach for optimizing the fuzzifier in fuzzy clustering.
    • The proposed method enhances clustering performance by adapting to dataset characteristics.
    • This work offers a valuable contribution to fuzzy clustering methodologies and their applications.