Geometric differential evolution for combinatorial and programs spaces

A. Moraglio, Julian Togelius, S. Silva

    Research output: Contribution to journalArticle

    Abstract

    Geometric differential evolution (GDE) is a recently introduced formal generalization of traditional differential evolution (DE) that can be used to derive specific differential evolution algorithms for both continuous and combinatorial spaces retaining the same geometric interpretation of the dynamics of the DE search across representations. In this article, we first review the theory behind the GDE algorithm, then, we use this framework to formally derive specific GDE for search spaces associated with binary strings, permutations, vectors of permutations and genetic programs. The resulting algorithms are representation-specific differential evolution algorithms searching the target spaces by acting directly on their underlying representations.We present experimental results for each of the new algorithms on a number of well-known problems comprising NK-landscapes, TSP, and Sudoku, for binary strings, permutations, and vectors of permutations.We also present results for the regression, artificial ant, parity, and multiplexer problems within the genetic programming domain. Experiments show that overall the new DE algorithms are competitive with well-tuned standard search algorithms.

    Original languageEnglish (US)
    Pages (from-to)591-624
    Number of pages34
    JournalEvolutionary Computation
    Volume21
    Issue number4
    DOIs
    StatePublished - Nov 2013

    Fingerprint

    Differential Evolution Algorithm
    Differential Evolution
    Permutation
    Strings
    Binary
    Geometric Algorithms
    Genetic Programming
    Search Space
    Parity
    Search Algorithm
    Regression
    Genetic programming
    Target
    Experimental Results
    Experiment

    Keywords

    • Combinatorial spaces
    • Differential evolution
    • Genetic programming
    • Principled design of search operators
    • Representations
    • Theory

    ASJC Scopus subject areas

    • Computational Mathematics

    Cite this

    Geometric differential evolution for combinatorial and programs spaces. / Moraglio, A.; Togelius, Julian; Silva, S.

    In: Evolutionary Computation, Vol. 21, No. 4, 11.2013, p. 591-624.

    Research output: Contribution to journalArticle

    Moraglio, A. ; Togelius, Julian ; Silva, S. / Geometric differential evolution for combinatorial and programs spaces. In: Evolutionary Computation. 2013 ; Vol. 21, No. 4. pp. 591-624.
    @article{2f65980ad73b4a8cb3ddc2557cfb6f9d,
    title = "Geometric differential evolution for combinatorial and programs spaces",
    abstract = "Geometric differential evolution (GDE) is a recently introduced formal generalization of traditional differential evolution (DE) that can be used to derive specific differential evolution algorithms for both continuous and combinatorial spaces retaining the same geometric interpretation of the dynamics of the DE search across representations. In this article, we first review the theory behind the GDE algorithm, then, we use this framework to formally derive specific GDE for search spaces associated with binary strings, permutations, vectors of permutations and genetic programs. The resulting algorithms are representation-specific differential evolution algorithms searching the target spaces by acting directly on their underlying representations.We present experimental results for each of the new algorithms on a number of well-known problems comprising NK-landscapes, TSP, and Sudoku, for binary strings, permutations, and vectors of permutations.We also present results for the regression, artificial ant, parity, and multiplexer problems within the genetic programming domain. Experiments show that overall the new DE algorithms are competitive with well-tuned standard search algorithms.",
    keywords = "Combinatorial spaces, Differential evolution, Genetic programming, Principled design of search operators, Representations, Theory",
    author = "A. Moraglio and Julian Togelius and S. Silva",
    year = "2013",
    month = "11",
    doi = "10.1162/EVCO_a_00099",
    language = "English (US)",
    volume = "21",
    pages = "591--624",
    journal = "Evolutionary Computation",
    issn = "1063-6560",
    publisher = "MIT Press Journals",
    number = "4",

    }

    TY - JOUR

    T1 - Geometric differential evolution for combinatorial and programs spaces

    AU - Moraglio, A.

    AU - Togelius, Julian

    AU - Silva, S.

    PY - 2013/11

    Y1 - 2013/11

    N2 - Geometric differential evolution (GDE) is a recently introduced formal generalization of traditional differential evolution (DE) that can be used to derive specific differential evolution algorithms for both continuous and combinatorial spaces retaining the same geometric interpretation of the dynamics of the DE search across representations. In this article, we first review the theory behind the GDE algorithm, then, we use this framework to formally derive specific GDE for search spaces associated with binary strings, permutations, vectors of permutations and genetic programs. The resulting algorithms are representation-specific differential evolution algorithms searching the target spaces by acting directly on their underlying representations.We present experimental results for each of the new algorithms on a number of well-known problems comprising NK-landscapes, TSP, and Sudoku, for binary strings, permutations, and vectors of permutations.We also present results for the regression, artificial ant, parity, and multiplexer problems within the genetic programming domain. Experiments show that overall the new DE algorithms are competitive with well-tuned standard search algorithms.

    AB - Geometric differential evolution (GDE) is a recently introduced formal generalization of traditional differential evolution (DE) that can be used to derive specific differential evolution algorithms for both continuous and combinatorial spaces retaining the same geometric interpretation of the dynamics of the DE search across representations. In this article, we first review the theory behind the GDE algorithm, then, we use this framework to formally derive specific GDE for search spaces associated with binary strings, permutations, vectors of permutations and genetic programs. The resulting algorithms are representation-specific differential evolution algorithms searching the target spaces by acting directly on their underlying representations.We present experimental results for each of the new algorithms on a number of well-known problems comprising NK-landscapes, TSP, and Sudoku, for binary strings, permutations, and vectors of permutations.We also present results for the regression, artificial ant, parity, and multiplexer problems within the genetic programming domain. Experiments show that overall the new DE algorithms are competitive with well-tuned standard search algorithms.

    KW - Combinatorial spaces

    KW - Differential evolution

    KW - Genetic programming

    KW - Principled design of search operators

    KW - Representations

    KW - Theory

    UR - http://www.scopus.com/inward/record.url?scp=84887157250&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84887157250&partnerID=8YFLogxK

    U2 - 10.1162/EVCO_a_00099

    DO - 10.1162/EVCO_a_00099

    M3 - Article

    VL - 21

    SP - 591

    EP - 624

    JO - Evolutionary Computation

    JF - Evolutionary Computation

    SN - 1063-6560

    IS - 4

    ER -