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

A path algorithm for the support vector domain description and its application to medical imaging.

Karl Sjöstrand1, Michael Sass Hansen, Henrik B Larsson

  • 1Informatics and Mathematical Modelling, Technical University of Denmark, Kgs. Lyngby, Denmark. kas@imm.dtu.dk

Medical Image Analysis
|September 8, 2007
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

Deep learning models built from PSMA PET of the primary tumor can predict synchronous and metachronous prostate cancer metastases.

PloS one·2026
Same author

Automated Imaging as an Adjunct to Serum and Clinical Biomarkers: A New Validated Prediction Tool for Metastatic Castration-Resistant Prostate Cancer.

Clinical cancer research : an official journal of the American Association for Cancer Research·2025
Same author

Quark Mass Dependence of Heavy Quark Diffusion Coefficient from Lattice QCD.

Physical review letters·2024
Same author

High-Entropy Oxides in the Mullite-Type Structure.

Chemistry of materials : a publication of the American Chemical Society·2023
Same author

Heavy Quark Diffusion from 2+1 Flavor Lattice QCD with 320 MeV Pion Mass.

Physical review letters·2023
Same author

Multimodal soft tissue markers for bridging high-resolution diagnostic imaging with therapeutic intervention.

Science advances·2020

This study introduces a new algorithm for support vector domain description (SVDD) that efficiently computes the entire regularization path. This method enhances outlier detection and model selection in complex datasets.

Area of Science:

  • Machine Learning
  • Data Mining
  • Pattern Recognition

Background:

  • Support Vector Domain Description (SVDD) is a one-class classification technique.
  • It estimates data distribution using a closed boundary to distinguish inliers from outliers.
  • Selecting the optimal regularization parameter is critical but computationally intensive.

Purpose of the Study:

  • To develop an efficient algorithm for computing the complete regularization path of SVDD.
  • To enable comprehensive model selection and data exploration through the regularization path.
  • To demonstrate the utility of the enhanced SVDD method in real-world applications.

Main Methods:

  • An algorithm is presented to compute all regularization path solutions for SVDD.
  • This is achieved with computational complexity similar to calculating a single SVDD solution.

Related Experiment Videos

  • The method explores the trade-off between boundary shape and outlier proportion.

    • Main Results:

    • The algorithm efficiently generates the entire regularization path for SVDD.
    • This provides a richer set of solutions than previously feasible.
    • Applications show improved data ordering and detection of myocardial ischemia.

    Conclusions:

    • The regularization path approach significantly enhances SVDD capabilities.
    • It facilitates better model selection and offers novel insights into data structure.
    • The method is effective for tasks like anatomical outline analysis and medical image segmentation.