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

Robust Bayesian clustering.

Cédric Archambeau1, Michel Verleysen

  • 1Machine Learning Group, Université catholique de Louvain, B-1348 Louvain-la-Neuve, Belgium. cedric.archambeau@uclouvain.be

Neural Networks : the Official Journal of the International Neural Network Society
|October 3, 2006
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

The Flow of Trust: A Visualization Framework to Externalize, Explore, and Explain Trust in ML Applications.

IEEE computer graphics and applications·2023
Same author

Fast Multiscale Neighbor Embedding.

IEEE transactions on neural networks and learning systems·2020
Same author

Dynamics of the perception and EEG signals triggered by tonic warm and cool stimulation.

PloS one·2020
Same author

Nonlinear Dimensionality Reduction With Missing Data Using Parametric Multiple Imputations.

IEEE transactions on neural networks and learning systems·2019
Same author

Reinforced Extreme Learning Machines for Fast Robust Regression in the Presence of Outliers.

IEEE transactions on cybernetics·2015
Same author

Latent IBP Compound Dirichlet Allocation.

IEEE transactions on pattern analysis and machine intelligence·2015

A novel variational Bayesian algorithm for Student-t mixture models enhances robust density estimation, clustering, and model selection. This approach improves upon Gaussian mixture models by better handling outliers for more reliable data analysis.

Area of Science:

  • Machine Learning
  • Statistical Modeling

Background:

  • Gaussian mixture models are widely used but sensitive to outliers.
  • Student-t distributions offer heavier tails, providing inherent robustness against outliers.

Purpose of the Study:

  • Introduce a new variational Bayesian learning algorithm for Student-t mixture models.
  • Develop a robust approach for density estimation, clustering, and model selection.

Main Methods:

  • Formalized a Bayesian Student-t mixture model as a latent variable model.
  • Avoided factorized approximations for posterior distributions of latent variables.
  • Incorporated correlations between unobserved random variables.

Main Results:

  • Achieved robust density estimation.

Related Experiment Videos

  • Enabled robust clustering.
  • Facilitated robust automatic model selection.
  • Obtained a tighter lower bound on the log-evidence.

    • Conclusions:

    • The proposed algorithm offers increased robustness compared to traditional methods.
    • The approach allows for more confident inference of model complexity (number of components).
    • This method advances robust Bayesian mixture modeling techniques.