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

Soft learning vector quantization and clustering algorithms based on non-Euclidean norms: single-norm algorithms.

Nicolaos B Karayiannis1, Mary M Randolph-Gips

  • 1Department of Electrical and Computer Engineering, University of Houston, Houston, TX 77204, USA. Karayiannis@UH.EDU

IEEE Transactions on Neural Networks
|March 25, 2005
PubMed
Summary

This study introduces novel soft clustering and Learning Vector Quantization (LVQ) algorithms using weighted norms for improved distance measurement. These methods offer competitive performance against computationally intensive algorithms, enhancing data analysis efficiency.

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 multistage system for the automated detection of epileptic seizures in neonatal electroencephalography.

Journal of clinical neurophysiology : official publication of the American Electroencephalographic Society·2009
Same author

Automated detection of videotaped neonatal seizures based on motion tracking methods.

Journal of clinical neurophysiology : official publication of the American Electroencephalographic Society·2006
Same author

Training reformulated radial basis function neural networks capable of identifying uncertainty in data classification.

IEEE transactions on neural networks·2006
Same author

Automated detection of videotaped neonatal seizures of epileptic origin.

Epilepsia·2006
Same author

Automated detection of videotaped neonatal seizures based on motion segmentation methods.

Clinical neurophysiology : official journal of the International Federation of Clinical Neurophysiology·2006
Same author

Detection of pseudosinusoidal epileptic seizure segments in the neonatal EEG by cascading a rule-based algorithm with a neural network.

IEEE transactions on bio-medical engineering·2006

Area of Science:

  • Machine Learning
  • Data Mining
  • Pattern Recognition

Background:

  • Traditional clustering and Learning Vector Quantization (LVQ) algorithms often rely on fixed distance metrics like the Euclidean norm.
  • These standard approaches may not optimally capture complex data structures, potentially limiting performance in diverse datasets.
  • Developing adaptive distance measures is crucial for enhancing the accuracy and robustness of clustering and prototype-based learning algorithms.

Purpose of the Study:

  • To develop novel soft clustering and LVQ algorithms utilizing a weighted norm for distance calculation.
  • To ensure the proposed algorithms are computationally efficient and comparable in speed to Euclidean norm-based methods.
  • To demonstrate the superior performance of the new algorithms compared to existing Euclidean norm-based approaches and computationally demanding non-Euclidean methods.

Related Experiment Videos

Main Methods:

  • Development of soft clustering and LVQ algorithms based on a weighted norm to measure feature vector-prototype distances.
  • Minimization of a reformulation function with a constraint on the generalized mean of norm weights.
  • Iterative computation of norm weights and prototypes directly from the data.
  • Error analysis to guide parameter selection based on feature variances.

Main Results:

  • The proposed algorithms are easy to implement and exhibit computational speed comparable to Euclidean norm-based clustering.
  • Experimental evaluation on four datasets shows consistent outperformance over Euclidean norm-based clustering algorithms.
  • The new algorithms demonstrate strong competitive performance against more computationally intensive non-Euclidean methods.

Conclusions:

  • The developed weighted norm-based soft clustering and LVQ algorithms provide a significant advancement over traditional Euclidean approaches.
  • These algorithms offer an effective and efficient alternative for pattern recognition tasks, especially with complex or varied datasets.
  • The findings suggest a promising direction for developing more adaptive and accurate machine learning algorithms.