Learning arithmetic read-once formulas

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

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

    Abstract

    A formula is read-once if each variable appears at most once in it. An arithmetic read-once formula is one in which the operators are addition, subtraction, multiplication, and division. We present polynomial time algorithms for exactly learning (or interpolating) arithmetic read-once formulas computing functions over a field. We present an algorithm that uses randomized membership queries (or substitutions) to identify such formulas over large finite fields and infinite fields. We also present a deterministic algorithm that uses equivalence queries as well as membership queries to identify arithmetic read-once formulas over small finite fields. We then non-constructively show the existence of deterministic membership query (interpolation) algorithms for arbitrary formulas over fields of characteristic 0 and for division-free formulas over large or infinite fields. Our algorithms assume we are able to efficiently perform arithmetic operations on field elements and compute square roots in the field. It is shown that the ability to compute square roots is necessary, in the sense that the problem of computing n-1 square roots in a field can be reduced to the problem of identifying an arithmetic formula over n variables in that field. Our equivalence queries are of a slightly non-standard form, in which counterexamples are required to not be inputs on which the formula evaluates to 0/0. This assumption is shown to be necessary for fields of size o(n/log n), for which it is shown that there is no polynomial time identification algorithm that uses just membership and standard equivalence queries.

    Original languageEnglish (US)
    Title of host publicationConference Proceedings of the Annual ACM Symposium on Theory of Computing
    Editors Anon
    PublisherPubl by ACM
    Pages370-381
    Number of pages12
    ISBN (Print)0897915119
    StatePublished - 1992
    EventProceedings of the 24th Annual ACM Symposium on the Theory of Computing - Victoria, BC, Can
    Duration: May 4 1992May 6 1992

    Other

    OtherProceedings of the 24th Annual ACM Symposium on the Theory of Computing
    CityVictoria, BC, Can
    Period5/4/925/6/92

    Fingerprint

    Polynomials
    Interpolation
    Substitution reactions

    ASJC Scopus subject areas

    • Software

    Cite this

    Bshouty, N. H., Hancock, T. R., & Hellerstein, L. (1992). Learning arithmetic read-once formulas. In Anon (Ed.), Conference Proceedings of the Annual ACM Symposium on Theory of Computing (pp. 370-381). Publ by ACM.

    Learning arithmetic read-once formulas. / Bshouty, Nader H.; Hancock, Thomas R.; Hellerstein, Lisa.

    Conference Proceedings of the Annual ACM Symposium on Theory of Computing. ed. / Anon. Publ by ACM, 1992. p. 370-381.

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

    Bshouty, NH, Hancock, TR & Hellerstein, L 1992, Learning arithmetic read-once formulas. in Anon (ed.), Conference Proceedings of the Annual ACM Symposium on Theory of Computing. Publ by ACM, pp. 370-381, Proceedings of the 24th Annual ACM Symposium on the Theory of Computing, Victoria, BC, Can, 5/4/92.
    Bshouty NH, Hancock TR, Hellerstein L. Learning arithmetic read-once formulas. In Anon, editor, Conference Proceedings of the Annual ACM Symposium on Theory of Computing. Publ by ACM. 1992. p. 370-381
    Bshouty, Nader H. ; Hancock, Thomas R. ; Hellerstein, Lisa. / Learning arithmetic read-once formulas. Conference Proceedings of the Annual ACM Symposium on Theory of Computing. editor / Anon. Publ by ACM, 1992. pp. 370-381
    @inproceedings{d6c6917264b24163990ce1669a289918,
    title = "Learning arithmetic read-once formulas",
    abstract = "A formula is read-once if each variable appears at most once in it. An arithmetic read-once formula is one in which the operators are addition, subtraction, multiplication, and division. We present polynomial time algorithms for exactly learning (or interpolating) arithmetic read-once formulas computing functions over a field. We present an algorithm that uses randomized membership queries (or substitutions) to identify such formulas over large finite fields and infinite fields. We also present a deterministic algorithm that uses equivalence queries as well as membership queries to identify arithmetic read-once formulas over small finite fields. We then non-constructively show the existence of deterministic membership query (interpolation) algorithms for arbitrary formulas over fields of characteristic 0 and for division-free formulas over large or infinite fields. Our algorithms assume we are able to efficiently perform arithmetic operations on field elements and compute square roots in the field. It is shown that the ability to compute square roots is necessary, in the sense that the problem of computing n-1 square roots in a field can be reduced to the problem of identifying an arithmetic formula over n variables in that field. Our equivalence queries are of a slightly non-standard form, in which counterexamples are required to not be inputs on which the formula evaluates to 0/0. This assumption is shown to be necessary for fields of size o(n/log n), for which it is shown that there is no polynomial time identification algorithm that uses just membership and standard equivalence queries.",
    author = "Bshouty, {Nader H.} and Hancock, {Thomas R.} and Lisa Hellerstein",
    year = "1992",
    language = "English (US)",
    isbn = "0897915119",
    pages = "370--381",
    editor = "Anon",
    booktitle = "Conference Proceedings of the Annual ACM Symposium on Theory of Computing",
    publisher = "Publ by ACM",

    }

    TY - GEN

    T1 - Learning arithmetic read-once formulas

    AU - Bshouty, Nader H.

    AU - Hancock, Thomas R.

    AU - Hellerstein, Lisa

    PY - 1992

    Y1 - 1992

    N2 - A formula is read-once if each variable appears at most once in it. An arithmetic read-once formula is one in which the operators are addition, subtraction, multiplication, and division. We present polynomial time algorithms for exactly learning (or interpolating) arithmetic read-once formulas computing functions over a field. We present an algorithm that uses randomized membership queries (or substitutions) to identify such formulas over large finite fields and infinite fields. We also present a deterministic algorithm that uses equivalence queries as well as membership queries to identify arithmetic read-once formulas over small finite fields. We then non-constructively show the existence of deterministic membership query (interpolation) algorithms for arbitrary formulas over fields of characteristic 0 and for division-free formulas over large or infinite fields. Our algorithms assume we are able to efficiently perform arithmetic operations on field elements and compute square roots in the field. It is shown that the ability to compute square roots is necessary, in the sense that the problem of computing n-1 square roots in a field can be reduced to the problem of identifying an arithmetic formula over n variables in that field. Our equivalence queries are of a slightly non-standard form, in which counterexamples are required to not be inputs on which the formula evaluates to 0/0. This assumption is shown to be necessary for fields of size o(n/log n), for which it is shown that there is no polynomial time identification algorithm that uses just membership and standard equivalence queries.

    AB - A formula is read-once if each variable appears at most once in it. An arithmetic read-once formula is one in which the operators are addition, subtraction, multiplication, and division. We present polynomial time algorithms for exactly learning (or interpolating) arithmetic read-once formulas computing functions over a field. We present an algorithm that uses randomized membership queries (or substitutions) to identify such formulas over large finite fields and infinite fields. We also present a deterministic algorithm that uses equivalence queries as well as membership queries to identify arithmetic read-once formulas over small finite fields. We then non-constructively show the existence of deterministic membership query (interpolation) algorithms for arbitrary formulas over fields of characteristic 0 and for division-free formulas over large or infinite fields. Our algorithms assume we are able to efficiently perform arithmetic operations on field elements and compute square roots in the field. It is shown that the ability to compute square roots is necessary, in the sense that the problem of computing n-1 square roots in a field can be reduced to the problem of identifying an arithmetic formula over n variables in that field. Our equivalence queries are of a slightly non-standard form, in which counterexamples are required to not be inputs on which the formula evaluates to 0/0. This assumption is shown to be necessary for fields of size o(n/log n), for which it is shown that there is no polynomial time identification algorithm that uses just membership and standard equivalence queries.

    UR - http://www.scopus.com/inward/record.url?scp=0026981387&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=0026981387&partnerID=8YFLogxK

    M3 - Conference contribution

    SN - 0897915119

    SP - 370

    EP - 381

    BT - Conference Proceedings of the Annual ACM Symposium on Theory of Computing

    A2 - Anon, null

    PB - Publ by ACM

    ER -