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

Bayesian feature and model selection for Gaussian mixture models.

Constantinos Constantinopoulos1, Michalis K Titsias, Aristidis Likas

  • 1Department of Computer Science, University of Ioannina, Ioannina GR 45110, Greece. ccostas@cs.uoi.gr

IEEE Transactions on Pattern Analysis and Machine Intelligence
|May 27, 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

AΚtransU-Net: Transformer-Equipped U-Net Model for Improved Actinic Keratosis Detection in Clinical Photography.

Diagnostics (Basel, Switzerland)·2025
Same author

The Promise of Semantic Segmentation in Detecting Actinic Keratosis Using Clinical Photography in the Wild.

Cancers·2023
Same author

Image Perceptual Similarity Metrics for the Assessment of Basal Cell Carcinoma.

Cancers·2023
Same author

Distance from Unimodality for the Assessment of Opinion Polarization.

Cognitive computation·2023
Same author

Scanning X-ray Fluorescence Data Analysis for the Identification of Byzantine Icons' Materials, Techniques, and State of Preservation: A Case Study.

Journal of imaging·2022
Same author

Context-Aware Automatic Sign Language Video Transcription in Psychiatric Interviews.

Sensors (Basel, Switzerland)·2022
Same journal

HardFlow: Hard-Constrained Sampling for Flow-Matching Models Via Trajectory Optimization.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Industrial Brain: Self-Evolving Neuro-Symbolic Autonomy with Causal Resilience for Cyber-Physical Systems.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Adaptive Hardness-Driven Dictionary Distillation for Incomplete Streaming View Clustering.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Mixture of Global and Local Experts with Diffusion Transformer for Controllable Face Generation.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Task-KV: Task-aware KV Cache Optimization via Semantic Differentiation of Attention Heads.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Achieving Text-based Person Retrieval with Any Granularity.

IEEE transactions on pattern analysis and machine intelligence·2026
See all related articles

This study introduces a Bayesian approach for training mixture models, simultaneously selecting features and the number of components. This method optimizes model complexity and feature relevance for improved data analysis.

Area of Science:

  • Machine Learning
  • Statistical Modeling
  • Computational Statistics

Background:

  • Mixture models are widely used for data clustering and density estimation.
  • Traditional methods often require pre-specifying the number of components and selecting relevant features.
  • Simultaneous feature and model selection remains a significant challenge in mixture model training.

Purpose of the Study:

  • To develop a unified Bayesian framework for mixture model training.
  • To simultaneously perform feature selection and determine the optimal number of mixture components.
  • To integrate feature saliency into the mixture model formulation.

Main Methods:

  • A Bayesian approach to mixture learning is employed.
  • A variational inference framework is utilized for optimization.

Related Experiment Videos

  • The method jointly optimizes the number of components, feature saliency, and model parameters.

    • Main Results:

    • The proposed method effectively performs simultaneous feature selection and model selection.
    • Experimental results on high-dimensional artificial and real data demonstrate the method's efficacy.
    • The Bayesian framework allows for principled estimation of the number of mixture components.

    Conclusions:

    • The developed Bayesian method offers an effective solution for mixture model training.
    • It addresses the challenges of feature selection and model selection in a unified manner.
    • The approach is robust and performs well on complex, high-dimensional datasets.