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 nearest prototype classification.

S Seo1, M Bode, K Obermayer

  • 1Dept. of Electr. Eng. and Comput. Sci., Tech. Univeristy of Berlin, Germany.

IEEE Transactions on Neural Networks
|February 2, 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

Mechanisms Tackling Salivary Gland Diseases with Extracellular Vesicle Therapies.

Journal of dental research·2025
Same author

Improved Measurements of Muonic Helium Ground-State Hyperfine Structure at a Near-Zero Magnetic Field.

Physical review letters·2024
Same author

Pre-Transplant Marital Status and Hematopoietic Cell Transplantation Outcomes

Current oncology (Toronto, Ont.)·2020
Same author

Release and toxicity comparison between industrial- and sunscreen-derived nano-ZnO particles.

International journal of environmental science and technology : IJEST·2020
Same author

Exogenous β-mannanase supplementation improved immunological and metabolic responses in lactating dairy cows.

Journal of dairy science·2019
Same author

Usefulness of TSH receptor antibodies as biomarkers for Graves' ophthalmopathy: a systematic review.

Journal of endocrinological investigation·2018
Same journal

Universal perceptron and DNA-like learning algorithm for binary neural networks: LSBF and PBF implementations.

IEEE transactions on neural networks·2013
Same journal

Guest editorial: special section on white box nonlinear prediction models.

IEEE transactions on neural networks·2011
Same journal

Data-based fault-tolerant control of high-speed trains with traction/braking notch nonlinearities and actuator failures.

IEEE transactions on neural networks·2011
Same journal

Guest editorial: special section on data-based control, modeling, and optimization.

IEEE transactions on neural networks·2011
Same journal

Neural network-based multiple robot simultaneous localization and mapping.

IEEE transactions on neural networks·2011
Same journal

Data-driven model-free adaptive control for a class of MIMO nonlinear discrete-time systems.

IEEE transactions on neural networks·2011
See all related articles

This study introduces an annealed Gaussian mixture method for nearest prototype classification, outperforming standard Learning Vector Quantization (LVQ) methods. Annealing key parameters significantly enhances classification accuracy and capability.

Area of Science:

  • Machine Learning
  • Pattern Recognition
  • Artificial Intelligence

Background:

  • Nearest prototype classifiers are essential for pattern recognition.
  • Standard Learning Vector Quantization (LVQ) methods have limitations.
  • A need exists for improved prototype-based classification algorithms.

Purpose of the Study:

  • To introduce a novel method for constructing nearest prototype classifiers.
  • To interpret the proposed method as an annealed version of LVQ.
  • To evaluate the performance of the new method against existing LVQ algorithms.

Main Methods:

  • The proposed method utilizes a Gaussian mixture ansatz.
  • It employs gradient descent on a cost function to minimize classification error.

Related Experiment Videos

  • Performance is assessed using toy datasets and an optical letter classification task.

    • Main Results:

    • Annealing the dispersion parameter of Gaussian kernels improves classification accuracy.
    • The proposed method achieves better results than standard LVQ (LVQ 2.1, LVQ 3) with equal prototypes.
    • Annealing the width parameter enhances classification capability.

    Conclusions:

    • The annealed Gaussian mixture approach offers superior performance to standard LVQ.
    • This principled method provides insights into the heuristic nature of LVQ.
    • The findings suggest potential for improved prototype-based classification systems.