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

Estimating the support of a high-dimensional distribution.

B Schölkopf1, J C Platt, J Shawe-Taylor

  • 1Microsoft Research Ltd, Cambridge CB2 3NH, U.K.

Neural Computation
|July 7, 2001
PubMed
Summary

This study introduces a novel algorithm for identifying simple data subsets by estimating a function using kernel expansion. This method extends support vector algorithms for unlabeled data, aiding in probability distribution analysis.

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

MRzero - Automated discovery of MRI sequences using supervised learning.

Magnetic resonance in medicine·2021
Same author

Achievement of Sustained Net Plasma Heating in a Fusion Experiment with the Optometrist Algorithm.

Scientific reports·2017
Same author

A cognitive brain-computer interface for patients with amyotrophic lateral sclerosis.

Progress in brain research·2016
Same author

Towards denoising XMCD movies of fast magnetization dynamics using extended Kalman filter.

Ultramicroscopy·2014
Same author

Identification of mRNA for endocannabinoid biosynthetic enzymes within hippocampal pyramidal cells and CA1 stratum radiatum interneuron subtypes using quantitative real-time polymerase chain reaction.

Neuroscience·2012
Same author

An online brain-computer interface based on shifting attention to concurrent streams of auditory stimuli.

Journal of neural engineering·2012

Area of Science:

  • Machine Learning
  • Statistical Learning Theory
  • Data Mining

Background:

  • Estimating simple subsets of input space from data is crucial for understanding probability distributions.
  • Existing methods may not efficiently handle unlabeled data for this task.

Purpose of the Study:

  • To propose a new algorithm for estimating a simple subset S from a dataset P.
  • To define a function f that is positive on S and negative on its complement.

Main Methods:

  • Utilizing a kernel expansion for the function f, regularized by controlling weight vector length.
  • Solving a quadratic programming problem via sequential optimization over input pattern pairs.
  • Extending the support vector algorithm to unlabeled data.

Related Experiment Videos

Main Results:

  • The proposed method effectively estimates the desired subset S.
  • Theoretical analysis confirms the statistical performance of the algorithm.
  • The algorithm provides a robust approach for unlabeled data.

Conclusions:

  • The developed algorithm offers a principled extension of support vector methods for unlabeled data.
  • This approach facilitates the estimation of simple subsets within complex probability distributions.
  • The method has potential applications in various data analysis and machine learning tasks.