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

Estimation of distribution algorithms with Kikuchi approximations.

Roberto Santana1

  • 1Institute of Cybernetics, Mathematics, and Physics, Calle 15, e/ C y D, Vedado, Cp-10400 Havana, Cuba. rsantana@si.ehu.es

Evolutionary Computation
|May 20, 2005
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

In silico neuronal morphology classification: A systematic review.

Neuroscience·2026
Same author

How Glucagon-Like Peptide-1 Medications Are Depicted in Instagram Posts Regarding Women's Health, Nontraditional Access, and Barriers to Access: Content Analysis.

Journal of medical Internet research·2025
Same author

Domain Adaptation-enhanced searchlight: enabling classification of brain states from visual perception to mental imagery.

Brain informatics·2025
Same author

Computational Tools for Neuronal Morphometric Analysis: A Systematic Search and Review.

Neuroinformatics·2024
Same author

Lobectomy in patients with differentiated thyroid cancer: experience of a Chilean tertiary center.

Endocrine·2024
Same author

Active surveillance is a feasible and safe strategy in selected patients with papillary thyroid cancer and suspicious cervical lymph nodes detected after thyroidectomy.

Archives of endocrinology and metabolism·2024
Same journal

Computing Optimal Populations for Binary Problems using Logic Minimization.

Evolutionary computation·2026
Same journal

Enhancing Generalization and Scalability for Multi-Objective Optimization with Population Pre-Training.

Evolutionary computation·2026
Same journal

XCS for Sequential Perceptual Aliasing in Multi-Step Decision Making.

Evolutionary computation·2026
Same journal

A dynamic multi-objective evolutionary algorithm using dual-space prediction and surrogate-based sampling.

Evolutionary computation·2026
Same journal

Adapting MOEA/D to CMA-ES for Dealing with Ill-conditioned Multiobjective Problems.

Evolutionary computation·2026
Same journal

Editorial of the Special Issue: Parallel Problem Solving from Nature PPSN 2024 Extended Versions of Best Paper Candidates.

Evolutionary computation·2026
See all related articles

This study introduces a new method for Estimation of Distribution Algorithms (EDAs) to learn more general probability factorizations. The novel Markov Network Estimation of Distribution Algorithm (MN-EDA) shows improved performance in optimizing functions with complex variable interactions.

Area of Science:

  • Artificial Intelligence
  • Machine Learning
  • Computational Statistics

Background:

  • Estimation of Distribution Algorithms (EDAs) rely on probability distribution estimation.
  • Current EDAs are limited by constrained factorizations from graphical models.
  • Expanding factorization capabilities is crucial for developing more robust EDAs.

Purpose of the Study:

  • To introduce a method for learning a more general class of probability factorizations for EDAs.
  • To develop a novel EDA that utilizes these generalized factorizations.
  • To evaluate the performance of the proposed algorithm against existing EDA methods.

Main Methods:

  • Combines Kikuchi approximation from statistical physics with novel graph decomposition techniques.
  • Introduces the Markov Network Estimation of Distribution Algorithm (MN-EDA).

Related Experiment Videos

  • Employs Gibbs Sampling (GS) for generating new data points within the MN-EDA framework.

    • Main Results:

    • MN-EDA learns a more general class of probability factorizations compared to traditional methods.
    • Empirical evaluations demonstrate MN-EDA's effectiveness.
    • MN-EDA outperforms Bayesian network-based EDAs on functions with strong variable interactions.

    Conclusions:

    • The proposed method significantly expands the factorization capabilities for EDAs.
    • MN-EDA offers a more robust approach to probability approximation in EDAs.
    • This advancement is particularly beneficial for optimization problems with complex dependencies.