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

Effects of kernel function on Nu support vector machines in extreme cases.

Kazushi Ikeda1

  • 1Graduate School of Informatics, Kyoto University, Kyoto 606-8501, Japan. kazushi@i.kyoto-u.ac.jp

IEEE Transactions on Neural Networks
|March 11, 2006
PubMed
Summary

Choosing a kernel function for support vector machines (SVMs) is challenging. This study analyzes v-SVM solutions for orthogonal and identical feature vectors, offering insights into kernel effects on generalization performance.

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

Theoretical analysis of the denoising autoencoder using Tweedie's formula.

Neural networks : the official journal of the International Neural Network Society·2026
Same author

The occurrence of a particular state is a predictor of successful travel consultation.

PloS one·2026
Same author

Herding as an emergent behaviour in harem groups of feral Garrano ponies.

Journal of the Royal Society, Interface·2026
Same author

Author Correction: Induction of functional tissue-engineered skeletal muscle constructs by defined electrical stimulation.

Scientific reports·2024
Same author

Time-series blood cytokine profiles correlate with treatment responses in triple-negative breast cancer patients.

British journal of cancer·2024
Same author

Bayesian traction force estimation using cell boundary-dependent force priors.

Biophysical journal·2023

Area of Science:

  • Machine Learning
  • Computational Statistics

Background:

  • Selecting appropriate kernel functions is crucial for Support Vector Machines (SVMs) performance.
  • Understanding kernel properties aids in optimizing generalization capabilities.

Purpose of the Study:

  • To investigate the behavior of v-SVM solutions with normalized feature vectors in extreme scenarios.
  • To analyze the impact of kernel choice on SVM generalization.

Main Methods:

  • Analysis of v-SVM solutions for normalized feature vectors.
  • Examination of two extreme cases: nearly orthogonal and nearly identical feature vectors.

Main Results:

  • For orthogonal vectors, v-SVM solutions approximate the center of gravity of the data.

Related Experiment Videos

  • For identical vectors, v-SVM solutions approach those of linear kernel SVMs.

    • Conclusions:

    • Extreme kernel analyses provide valuable insights into kernel function influence on generalization.
    • Understanding these properties can guide better kernel selection in practical SVM applications.