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

Learning mixture models with the regularized latent maximum entropy principle.

Shaojun Wang1, Dale Schuurmans, Fuchun Peng

  • 1Department of Computing Science, University of Alberta, Alberta T6G 2E8, Canada.

IEEE Transactions on Neural Networks
|October 6, 2004
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

Chiglitazar in combination with anti-inflammatory and hepatoprotective therapy for the treatment of MASH associated with T2DM: a prospective, multicentre, randomised, double-blind, placebo-controlled study protocol.

BMJ open·2026
Same author

NPRL2 restricts porcine reproductive and respiratory syndrome virus replication by targeting viral Nsp1α for autophagic degradation.

Veterinary research·2026
Same author

Predictive Factors for Ciliary Block after Cataract and Goniosynechialysis Surgery in Primary Angle-Closure Disease.

Ophthalmology. Glaucoma·2026
Same author

Longitudinal relation of family dysfunction to young adults' social anxiety: Sex and baseline respiratory sinus arrhythmia matter.

Biological psychology·2025
Same author

Unidirectional Giant Exciton Emission into a Photonic Waveguide.

Physical review letters·2025
Same author

Phytochemicals as modulators of Astrocytes in Alzheimer's disease: A therapeutic perspective.

Phytomedicine : international journal of phytotherapy and phytopharmacology·2025
Same journal

Universal perceptron and DNA-like learning algorithm for binary neural networks: LSBF and PBF implementations.

IEEE transactions on neural networks·2013
Same journal

Guest editorial: special section on white box nonlinear prediction models.

IEEE transactions on neural networks·2011
Same journal

Data-based fault-tolerant control of high-speed trains with traction/braking notch nonlinearities and actuator failures.

IEEE transactions on neural networks·2011
Same journal

Guest editorial: special section on data-based control, modeling, and optimization.

IEEE transactions on neural networks·2011
Same journal

Neural network-based multiple robot simultaneous localization and mapping.

IEEE transactions on neural networks·2011
Same journal

Data-driven model-free adaptive control for a class of MIMO nonlinear discrete-time systems.

IEEE transactions on neural networks·2011
See all related articles

This study introduces the latent maximum entropy principle (LME) for mixture model estimation. Regularized LME (RLME) offers improved accuracy over traditional methods, especially with limited data.

Area of Science:

  • Statistics
  • Machine Learning
  • Computational Statistics

Background:

  • Mixture models are widely used for data analysis.
  • Existing estimation methods like maximum likelihood and expectation-maximization have limitations, particularly with small datasets.
  • Novel inference principles are needed for robust model estimation.

Purpose of the Study:

  • To introduce and demonstrate the latent maximum entropy principle (LME) for mixture model estimation.
  • To develop new algorithms and robust variants of the expectation-maximization (EM) algorithm based on LME.
  • To evaluate the performance of LME and its regularized version (RLME) against established methods.

Main Methods:

  • Development of the latent maximum entropy principle (LME).

Related Experiment Videos

  • Derivation of new mixture model estimation algorithms based on LME.
  • Formulation of regularized LME (RLME).
  • Comparison with maximum likelihood and maximum a posteriori (MAP) estimation.

    • Main Results:

    • LME provides a novel framework for mixture model estimation.
    • RLME demonstrates effectiveness and robustness in estimating mixture models.
    • RLME generally outperforms plain LME, maximum likelihood, and MAP estimation.
    • LME-based methods show particular advantages when inferring latent variable models from small data samples.

    Conclusions:

    • The latent maximum entropy principle (LME) offers a powerful new approach to mixture model estimation.
    • Regularized LME (RLME) provides superior performance, especially in data-scarce scenarios.
    • LME-based algorithms represent a significant advancement over traditional estimation techniques for latent variable models.