On PAC learning algorithms for rich Boolean function classes

Lisa Hellerstein, Rocco A. Servedio

    Research output: Contribution to journalArticle

    Abstract

    We give an overview of the fastest known algorithms for learning various expressive classes of Boolean functions in the Probably Approximately Correct (PAC) learning model. In addition to surveying previously known results, we use existing techniques to give the first known subexponential-time algorithms for PAC learning two natural and expressive classes of Boolean functions: sparse polynomial threshold functions over the Boolean cube {0, 1}n and sparse GF2 polynomials over {0, 1}n.

    Original languageEnglish (US)
    Pages (from-to)66-76
    Number of pages11
    JournalTheoretical Computer Science
    Volume384
    Issue number1
    DOIs
    StatePublished - Sep 24 2007

    Fingerprint

    Boolean functions
    Boolean Functions
    Learning algorithms
    Sparse Polynomials
    Learning Algorithm
    Polynomials
    Surveying
    Threshold Function
    Polynomial function
    Fast Algorithm
    Regular hexahedron
    Learning
    Class
    Model

    Keywords

    • Computational learning theory
    • PAC learning
    • Polynomial threshold function

    ASJC Scopus subject areas

    • Computational Theory and Mathematics

    Cite this

    On PAC learning algorithms for rich Boolean function classes. / Hellerstein, Lisa; Servedio, Rocco A.

    In: Theoretical Computer Science, Vol. 384, No. 1, 24.09.2007, p. 66-76.

    Research output: Contribution to journalArticle

    Hellerstein, Lisa ; Servedio, Rocco A. / On PAC learning algorithms for rich Boolean function classes. In: Theoretical Computer Science. 2007 ; Vol. 384, No. 1. pp. 66-76.
    @article{1a69cc02fb5449f4a1380fcb69300cb2,
    title = "On PAC learning algorithms for rich Boolean function classes",
    abstract = "We give an overview of the fastest known algorithms for learning various expressive classes of Boolean functions in the Probably Approximately Correct (PAC) learning model. In addition to surveying previously known results, we use existing techniques to give the first known subexponential-time algorithms for PAC learning two natural and expressive classes of Boolean functions: sparse polynomial threshold functions over the Boolean cube {0, 1}n and sparse GF2 polynomials over {0, 1}n.",
    keywords = "Computational learning theory, PAC learning, Polynomial threshold function",
    author = "Lisa Hellerstein and Servedio, {Rocco A.}",
    year = "2007",
    month = "9",
    day = "24",
    doi = "10.1016/j.tcs.2007.05.018",
    language = "English (US)",
    volume = "384",
    pages = "66--76",
    journal = "Theoretical Computer Science",
    issn = "0304-3975",
    publisher = "Elsevier",
    number = "1",

    }

    TY - JOUR

    T1 - On PAC learning algorithms for rich Boolean function classes

    AU - Hellerstein, Lisa

    AU - Servedio, Rocco A.

    PY - 2007/9/24

    Y1 - 2007/9/24

    N2 - We give an overview of the fastest known algorithms for learning various expressive classes of Boolean functions in the Probably Approximately Correct (PAC) learning model. In addition to surveying previously known results, we use existing techniques to give the first known subexponential-time algorithms for PAC learning two natural and expressive classes of Boolean functions: sparse polynomial threshold functions over the Boolean cube {0, 1}n and sparse GF2 polynomials over {0, 1}n.

    AB - We give an overview of the fastest known algorithms for learning various expressive classes of Boolean functions in the Probably Approximately Correct (PAC) learning model. In addition to surveying previously known results, we use existing techniques to give the first known subexponential-time algorithms for PAC learning two natural and expressive classes of Boolean functions: sparse polynomial threshold functions over the Boolean cube {0, 1}n and sparse GF2 polynomials over {0, 1}n.

    KW - Computational learning theory

    KW - PAC learning

    KW - Polynomial threshold function

    UR - http://www.scopus.com/inward/record.url?scp=34548203510&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=34548203510&partnerID=8YFLogxK

    U2 - 10.1016/j.tcs.2007.05.018

    DO - 10.1016/j.tcs.2007.05.018

    M3 - Article

    AN - SCOPUS:34548203510

    VL - 384

    SP - 66

    EP - 76

    JO - Theoretical Computer Science

    JF - Theoretical Computer Science

    SN - 0304-3975

    IS - 1

    ER -