Scalability and efficiency of genetic algorithms for geometrical applications

S.F. Dijk, van, D. Thierens, M. Berg, de

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

    5 Citations (Scopus)

    Abstract

    We study the scalability and efficiency of a GA that we developed earlier to solve the practical cartographic problem of labeling a map with point features. We argue that the special characteristics of our GA make that it fits in well with theoretical models predicting the optimal population size (the Gambler’s Ruin model) and the number of generations until convergence. We then verify these predictions experimentally. It turns out that our algorithm indeed performs according to the theory, leading to a scale-up for the total amount of computational effort that is linear in the problem size.
    Original languageEnglish
    Title of host publicationParallel Problem Solving from Nature (Proceedings 6th International Conference, PPSN-VI, Paris, France, September 18-20, 2000)
    EditorsM. Schoenauer
    Place of PublicationBerlin
    PublisherSpringer
    Pages683-692
    ISBN (Print)3-540-41056-2
    DOIs
    Publication statusPublished - 2000

    Publication series

    NameLecture Notes in Computer Science
    Volume1917
    ISSN (Print)0302-9743

    Fingerprint Dive into the research topics of 'Scalability and efficiency of genetic algorithms for geometrical applications'. Together they form a unique fingerprint.

    Cite this