PhysicsGP: A genetic programming approach to event selection

Kyle Cranmer, R. Sean Bowman

    Research output: Contribution to journalArticle

    Abstract

    We present a novel multivariate classification technique based on Genetic Programming. The technique is distinct from Genetic Algorithms and offers several advantages compared to Neural Networks and Support Vector Machines. The technique optimizes a set of human-readable classifiers with respect to some user-defined performance measure. We calculate the Vapnik-Chervonenkis dimension of this class of learning machines and consider a practical example: the search for the Standard Model Higgs Boson at the LHC. The resulting classifier is very fast to evaluate, human-readable, and easily portable. The software may be downloaded at: http://cern.ch/~cranmer/PhysicsGP.html.

    Original languageEnglish (US)
    Pages (from-to)165-167
    Number of pages3
    JournalComputer Physics Communications
    Volume167
    Issue number3
    DOIs
    StatePublished - May 1 2005

    Fingerprint

    Genetic programming
    classifiers
    programming
    Classifiers
    Bosons
    machine learning
    Higgs bosons
    genetic algorithms
    Support vector machines
    Learning systems
    Genetic algorithms
    Neural networks
    computer programs

    Keywords

    • Classification
    • Genetic algorithms
    • Genetic Programming
    • Neural networks
    • Support vector machines
    • Triggering
    • VC dimension

    ASJC Scopus subject areas

    • Computer Science Applications
    • Physics and Astronomy(all)

    Cite this

    PhysicsGP : A genetic programming approach to event selection. / Cranmer, Kyle; Bowman, R. Sean.

    In: Computer Physics Communications, Vol. 167, No. 3, 01.05.2005, p. 165-167.

    Research output: Contribution to journalArticle

    Cranmer, Kyle ; Bowman, R. Sean. / PhysicsGP : A genetic programming approach to event selection. In: Computer Physics Communications. 2005 ; Vol. 167, No. 3. pp. 165-167.
    @article{bec1b84136e1416980b7e298b1ac5fbb,
    title = "PhysicsGP: A genetic programming approach to event selection",
    abstract = "We present a novel multivariate classification technique based on Genetic Programming. The technique is distinct from Genetic Algorithms and offers several advantages compared to Neural Networks and Support Vector Machines. The technique optimizes a set of human-readable classifiers with respect to some user-defined performance measure. We calculate the Vapnik-Chervonenkis dimension of this class of learning machines and consider a practical example: the search for the Standard Model Higgs Boson at the LHC. The resulting classifier is very fast to evaluate, human-readable, and easily portable. The software may be downloaded at: http://cern.ch/~cranmer/PhysicsGP.html.",
    keywords = "Classification, Genetic algorithms, Genetic Programming, Neural networks, Support vector machines, Triggering, VC dimension",
    author = "Kyle Cranmer and Bowman, {R. Sean}",
    year = "2005",
    month = "5",
    day = "1",
    doi = "10.1016/j.cpc.2004.12.006",
    language = "English (US)",
    volume = "167",
    pages = "165--167",
    journal = "Computer Physics Communications",
    issn = "0010-4655",
    publisher = "Elsevier",
    number = "3",

    }

    TY - JOUR

    T1 - PhysicsGP

    T2 - A genetic programming approach to event selection

    AU - Cranmer, Kyle

    AU - Bowman, R. Sean

    PY - 2005/5/1

    Y1 - 2005/5/1

    N2 - We present a novel multivariate classification technique based on Genetic Programming. The technique is distinct from Genetic Algorithms and offers several advantages compared to Neural Networks and Support Vector Machines. The technique optimizes a set of human-readable classifiers with respect to some user-defined performance measure. We calculate the Vapnik-Chervonenkis dimension of this class of learning machines and consider a practical example: the search for the Standard Model Higgs Boson at the LHC. The resulting classifier is very fast to evaluate, human-readable, and easily portable. The software may be downloaded at: http://cern.ch/~cranmer/PhysicsGP.html.

    AB - We present a novel multivariate classification technique based on Genetic Programming. The technique is distinct from Genetic Algorithms and offers several advantages compared to Neural Networks and Support Vector Machines. The technique optimizes a set of human-readable classifiers with respect to some user-defined performance measure. We calculate the Vapnik-Chervonenkis dimension of this class of learning machines and consider a practical example: the search for the Standard Model Higgs Boson at the LHC. The resulting classifier is very fast to evaluate, human-readable, and easily portable. The software may be downloaded at: http://cern.ch/~cranmer/PhysicsGP.html.

    KW - Classification

    KW - Genetic algorithms

    KW - Genetic Programming

    KW - Neural networks

    KW - Support vector machines

    KW - Triggering

    KW - VC dimension

    UR - http://www.scopus.com/inward/record.url?scp=17044413299&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=17044413299&partnerID=8YFLogxK

    U2 - 10.1016/j.cpc.2004.12.006

    DO - 10.1016/j.cpc.2004.12.006

    M3 - Article

    VL - 167

    SP - 165

    EP - 167

    JO - Computer Physics Communications

    JF - Computer Physics Communications

    SN - 0010-4655

    IS - 3

    ER -