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

FDA - a scalable evolutionary algorithm for the optimization of additively decomposed functions.

H Muehlenbein1, T Mahnig

  • 1Theoretical Foundation GMD Lab., Real World Computing Partnership, GMD FZ Informationstechnik, 53754 St. Augustin, Germany. muehlenbein@gmd.de

Evolutionary Computation
|December 1, 1999
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 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

The Factorized Distribution Algorithm (FDA) optimizes complex problems by estimating distributions. This evolutionary algorithm efficiently solves difficult functions using factorized distributions, outperforming standard genetic algorithms.

Area of Science:

  • Evolutionary Computation
  • Artificial Intelligence
  • Optimization Algorithms

Background:

  • Traditional evolutionary algorithms face challenges with high-dimensional problems due to computational complexity.
  • Estimating distributions for discrete variables requires a large number of parameters, hindering scalability.
  • Additively Decomposed Functions (ADFs) offer a structure that can be exploited for more efficient computation.

Purpose of the Study:

  • To investigate the theoretical and numerical scaling of the Factorized Distribution Algorithm (FDA).
  • To demonstrate FDA's effectiveness on difficult functions, particularly those with specific structures like chains or trees.
  • To extend FDA using Bayesian network principles to create LFDA for approximate factorizations.

Main Methods:

Related Experiment Videos

  • FDA combines mutation and recombination by estimating a distribution from selected points.
  • Factorization of distributions into conditional and marginal distributions is key to FDA's efficiency.
  • The study analyzes FDA's performance on various structures, including exact and approximate factorizations.
  • LFDA is developed by incorporating Bayesian network insights for data-driven approximate factorization.

    • Main Results:

    • FDA demonstrates efficient scaling, solving difficult functions on chain or tree structures in approximately O(n√n) operations.
    • FDA successfully optimizes functions using both exact and approximate factorizations, as shown on circle and grid structures.
    • LFDA provides an alternative approach to approximate factorization using only data, without needing the ADF structure.
    • The scaling of LFDA is compared to that of FDA, highlighting different performance characteristics.

    Conclusions:

    • FDA offers a significant advantage over standard genetic algorithms for optimizing complex, additively decomposable functions.
    • The algorithm's efficiency is tied to the underlying function structure and factorization approach.
    • FDA's adaptability to approximate factorizations and extension to LFDA broaden its applicability in optimization problems.