Using self-learning and automatic tuning to improve the performance of sexual genetic algorithms for constraint satisfaction problems

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    Currently the parameters in a constraint solver are often selected by hand by experts in the field; these parameters might include the level of preprocessing to be used and the variable ordering heuristic. The efficient and automatic choice of a preprocessing level for a constraint solver is a step towards making constraint programming a more widely accessible technology. Self-learning sexual genetic algorithms are a new approach combining a self-learning mechanism with sexual genetic algorithms in order to suggest or predict a suitable solver configuration for large scale problems by learning from the same class of small scale problems. In this paper, Self-learning Sexual genetic algorithms are applied to create an automatic solver configuration mechanism for solving various constraint problems. The starting population of self-learning sexual genetic algorithms will be trained through experience on small instances. The experiments in this paper are a proof-of-concept for the idea of combining sexual genetic algorithms with a self-learning strategy to aid in parameter selection for constraint programming.

    Original languageEnglish
    Title of host publication2013 Imperial College Computing Student Workshop (ICCSW'13)
    EditorsAndrew V. Jones, Nicholas Ng
    Place of PublicationSaarbrücken
    PublisherSchloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH
    Pages128-135
    Number of pages8
    ISBN (Print)9783939897637
    DOIs
    Publication statusPublished - 2013
    Event2013 Imperial College Computing Student Workshop - Department of Computing located inside the South Kensington Campus at Imperial College London, London, United Kingdom
    Duration: 26 Sep 201327 Sep 2013
    http://iccsw.doc.ic.ac.uk/2013/

    Publication series

    NameOASICS
    Volume35
    ISSN (Print)2190-6807

    Workshop

    Workshop2013 Imperial College Computing Student Workshop
    Abbreviated titleICCSW'13
    CountryUnited Kingdom
    CityLondon
    Period26/09/1327/09/13
    Internet address

    Fingerprint

    Constraint satisfaction problems
    Self-learning
    Constraint Satisfaction Problem
    genetic algorithm
    Tuning
    Genetic algorithms
    learning
    Genetic Algorithm
    Learning algorithms
    Learning Algorithm
    Constraint Programming
    performance
    Preprocessing
    programming
    Constraint Solving
    Configuration
    Learning Strategies
    Parameter Selection
    Large-scale Problems
    learning strategy

    Keywords

    • Constraint programming
    • Parameter tuning
    • Self-learning genetic algorithm
    • Sexual genetic algorithm

    Cite this

    Xu, H., Petrie, K., & Murray, I. (2013). Using self-learning and automatic tuning to improve the performance of sexual genetic algorithms for constraint satisfaction problems. In A. V. Jones, & N. Ng (Eds.), 2013 Imperial College Computing Student Workshop (ICCSW'13) (pp. 128-135). (OASICS ; Vol. 35). Saarbrücken: Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH. https://doi.org/10.4230/OASIcs.ICCSW.2013.128
    Xu, Hu ; Petrie, Karen ; Murray, Iain. / Using self-learning and automatic tuning to improve the performance of sexual genetic algorithms for constraint satisfaction problems. 2013 Imperial College Computing Student Workshop (ICCSW'13). editor / Andrew V. Jones ; Nicholas Ng. Saarbrücken : Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, 2013. pp. 128-135 (OASICS ).
    @inproceedings{127e32243e764d899621be1b4d238d2d,
    title = "Using self-learning and automatic tuning to improve the performance of sexual genetic algorithms for constraint satisfaction problems",
    abstract = "Currently the parameters in a constraint solver are often selected by hand by experts in the field; these parameters might include the level of preprocessing to be used and the variable ordering heuristic. The efficient and automatic choice of a preprocessing level for a constraint solver is a step towards making constraint programming a more widely accessible technology. Self-learning sexual genetic algorithms are a new approach combining a self-learning mechanism with sexual genetic algorithms in order to suggest or predict a suitable solver configuration for large scale problems by learning from the same class of small scale problems. In this paper, Self-learning Sexual genetic algorithms are applied to create an automatic solver configuration mechanism for solving various constraint problems. The starting population of self-learning sexual genetic algorithms will be trained through experience on small instances. The experiments in this paper are a proof-of-concept for the idea of combining sexual genetic algorithms with a self-learning strategy to aid in parameter selection for constraint programming.",
    keywords = "Constraint programming, Parameter tuning, Self-learning genetic algorithm, Sexual genetic algorithm",
    author = "Hu Xu and Karen Petrie and Iain Murray",
    year = "2013",
    doi = "10.4230/OASIcs.ICCSW.2013.128",
    language = "English",
    isbn = "9783939897637",
    series = "OASICS",
    publisher = "Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH",
    pages = "128--135",
    editor = "Jones, {Andrew V. } and Nicholas Ng",
    booktitle = "2013 Imperial College Computing Student Workshop (ICCSW'13)",
    address = "Germany",

    }

    Xu, H, Petrie, K & Murray, I 2013, Using self-learning and automatic tuning to improve the performance of sexual genetic algorithms for constraint satisfaction problems. in AV Jones & N Ng (eds), 2013 Imperial College Computing Student Workshop (ICCSW'13). OASICS , vol. 35, Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, Saarbrücken, pp. 128-135, 2013 Imperial College Computing Student Workshop, London, United Kingdom, 26/09/13. https://doi.org/10.4230/OASIcs.ICCSW.2013.128

    Using self-learning and automatic tuning to improve the performance of sexual genetic algorithms for constraint satisfaction problems. / Xu, Hu; Petrie, Karen; Murray, Iain.

    2013 Imperial College Computing Student Workshop (ICCSW'13). ed. / Andrew V. Jones; Nicholas Ng. Saarbrücken : Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, 2013. p. 128-135 (OASICS ; Vol. 35).

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    TY - GEN

    T1 - Using self-learning and automatic tuning to improve the performance of sexual genetic algorithms for constraint satisfaction problems

    AU - Xu, Hu

    AU - Petrie, Karen

    AU - Murray, Iain

    PY - 2013

    Y1 - 2013

    N2 - Currently the parameters in a constraint solver are often selected by hand by experts in the field; these parameters might include the level of preprocessing to be used and the variable ordering heuristic. The efficient and automatic choice of a preprocessing level for a constraint solver is a step towards making constraint programming a more widely accessible technology. Self-learning sexual genetic algorithms are a new approach combining a self-learning mechanism with sexual genetic algorithms in order to suggest or predict a suitable solver configuration for large scale problems by learning from the same class of small scale problems. In this paper, Self-learning Sexual genetic algorithms are applied to create an automatic solver configuration mechanism for solving various constraint problems. The starting population of self-learning sexual genetic algorithms will be trained through experience on small instances. The experiments in this paper are a proof-of-concept for the idea of combining sexual genetic algorithms with a self-learning strategy to aid in parameter selection for constraint programming.

    AB - Currently the parameters in a constraint solver are often selected by hand by experts in the field; these parameters might include the level of preprocessing to be used and the variable ordering heuristic. The efficient and automatic choice of a preprocessing level for a constraint solver is a step towards making constraint programming a more widely accessible technology. Self-learning sexual genetic algorithms are a new approach combining a self-learning mechanism with sexual genetic algorithms in order to suggest or predict a suitable solver configuration for large scale problems by learning from the same class of small scale problems. In this paper, Self-learning Sexual genetic algorithms are applied to create an automatic solver configuration mechanism for solving various constraint problems. The starting population of self-learning sexual genetic algorithms will be trained through experience on small instances. The experiments in this paper are a proof-of-concept for the idea of combining sexual genetic algorithms with a self-learning strategy to aid in parameter selection for constraint programming.

    KW - Constraint programming

    KW - Parameter tuning

    KW - Self-learning genetic algorithm

    KW - Sexual genetic algorithm

    UR - http://www.scopus.com/inward/record.url?scp=84905827371&partnerID=8YFLogxK

    U2 - 10.4230/OASIcs.ICCSW.2013.128

    DO - 10.4230/OASIcs.ICCSW.2013.128

    M3 - Conference contribution

    AN - SCOPUS:84905827371

    SN - 9783939897637

    T3 - OASICS

    SP - 128

    EP - 135

    BT - 2013 Imperial College Computing Student Workshop (ICCSW'13)

    A2 - Jones, Andrew V.

    A2 - Ng, Nicholas

    PB - Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH

    CY - Saarbrücken

    ER -

    Xu H, Petrie K, Murray I. Using self-learning and automatic tuning to improve the performance of sexual genetic algorithms for constraint satisfaction problems. In Jones AV, Ng N, editors, 2013 Imperial College Computing Student Workshop (ICCSW'13). Saarbrücken: Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH. 2013. p. 128-135. (OASICS ). https://doi.org/10.4230/OASIcs.ICCSW.2013.128