An algorithm to learn read-once threshold formulas, and transformations between learning models

Nader H. Bshouty, Thomas R. Hancock, Lisa Hellerstein, Marek Karpinski

    Research output: Contribution to journalArticle

    Abstract

    We present a membership query (i.e. black box interpolation) algorithm for exactly identifying the class of read-once formulas over the basis of Boolean threshold functions. We also present a catalogue of generic transformations that can be used to convert an algorithm in one learning model into an algorithm in a different model.

    Original languageEnglish (US)
    Pages (from-to)37-61
    Number of pages25
    JournalComputational Complexity
    Volume4
    Issue number1
    DOIs
    StatePublished - Mar 1994

    Fingerprint

    Threshold Function
    Boolean Functions
    Black Box
    Convert
    Interpolation
    Interpolate
    Model
    Query
    Learning
    Class

    Keywords

    • 68Q20
    • 68T05
    • equivalence queries
    • justifying assignments
    • learning algorithms
    • learning models
    • membership queries
    • read-once threshold formulas
    • Subject classifications
    • transformations

    ASJC Scopus subject areas

    • Theoretical Computer Science
    • Computational Mathematics
    • Mathematics(all)
    • Computational Theory and Mathematics

    Cite this

    An algorithm to learn read-once threshold formulas, and transformations between learning models. / Bshouty, Nader H.; Hancock, Thomas R.; Hellerstein, Lisa; Karpinski, Marek.

    In: Computational Complexity, Vol. 4, No. 1, 03.1994, p. 37-61.

    Research output: Contribution to journalArticle

    Bshouty, Nader H. ; Hancock, Thomas R. ; Hellerstein, Lisa ; Karpinski, Marek. / An algorithm to learn read-once threshold formulas, and transformations between learning models. In: Computational Complexity. 1994 ; Vol. 4, No. 1. pp. 37-61.
    @article{e722acfdd4f848ef954acfd4d9e916c4,
    title = "An algorithm to learn read-once threshold formulas, and transformations between learning models",
    abstract = "We present a membership query (i.e. black box interpolation) algorithm for exactly identifying the class of read-once formulas over the basis of Boolean threshold functions. We also present a catalogue of generic transformations that can be used to convert an algorithm in one learning model into an algorithm in a different model.",
    keywords = "68Q20, 68T05, equivalence queries, justifying assignments, learning algorithms, learning models, membership queries, read-once threshold formulas, Subject classifications, transformations",
    author = "Bshouty, {Nader H.} and Hancock, {Thomas R.} and Lisa Hellerstein and Marek Karpinski",
    year = "1994",
    month = "3",
    doi = "10.1007/BF01205054",
    language = "English (US)",
    volume = "4",
    pages = "37--61",
    journal = "Computational Complexity",
    issn = "1016-3328",
    publisher = "Birkhauser Verlag Basel",
    number = "1",

    }

    TY - JOUR

    T1 - An algorithm to learn read-once threshold formulas, and transformations between learning models

    AU - Bshouty, Nader H.

    AU - Hancock, Thomas R.

    AU - Hellerstein, Lisa

    AU - Karpinski, Marek

    PY - 1994/3

    Y1 - 1994/3

    N2 - We present a membership query (i.e. black box interpolation) algorithm for exactly identifying the class of read-once formulas over the basis of Boolean threshold functions. We also present a catalogue of generic transformations that can be used to convert an algorithm in one learning model into an algorithm in a different model.

    AB - We present a membership query (i.e. black box interpolation) algorithm for exactly identifying the class of read-once formulas over the basis of Boolean threshold functions. We also present a catalogue of generic transformations that can be used to convert an algorithm in one learning model into an algorithm in a different model.

    KW - 68Q20

    KW - 68T05

    KW - equivalence queries

    KW - justifying assignments

    KW - learning algorithms

    KW - learning models

    KW - membership queries

    KW - read-once threshold formulas

    KW - Subject classifications

    KW - transformations

    UR - http://www.scopus.com/inward/record.url?scp=0008165967&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=0008165967&partnerID=8YFLogxK

    U2 - 10.1007/BF01205054

    DO - 10.1007/BF01205054

    M3 - Article

    AN - SCOPUS:0008165967

    VL - 4

    SP - 37

    EP - 61

    JO - Computational Complexity

    JF - Computational Complexity

    SN - 1016-3328

    IS - 1

    ER -