Budget feasible mechanisms for experimental design

Thibaut Horel, Stratis Ioannidis, Shanmugavelayutham Muthukrishnan

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

    Abstract

    We present a deterministic, polynomial time, budget feasible mechanism scheme, that is approximately truthful and yields a constant (≈ 12.98) factor approximation for the Experimental Design Problem (EDP). By applying previous work on budget feasible mechanisms with a submodular objective, one could only have derived either an exponential time deterministic mechanism or a randomized polynomial time mechanism. We also establish that no truthful, budget-feasible mechanism is possible within a factor 2 approximation, and show how to generalize our approach to a wide class of learning problems, beyond linear regression.

    Original languageEnglish (US)
    Title of host publicationLATIN 2014
    Subtitle of host publicationTheoretical Informatics - 11th Latin American Symposium, Proceedings
    PublisherSpringer-Verlag
    Pages719-730
    Number of pages12
    ISBN (Print)9783642544224
    DOIs
    StatePublished - Jan 1 2014
    Event11th Latin American Theoretical Informatics Symposium, LATIN 2014 - Montevideo, Uruguay
    Duration: Mar 31 2014Apr 4 2014

    Publication series

    NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    Volume8392 LNCS
    ISSN (Print)0302-9743
    ISSN (Electronic)1611-3349

    Conference

    Conference11th Latin American Theoretical Informatics Symposium, LATIN 2014
    CountryUruguay
    CityMontevideo
    Period3/31/144/4/14

    Fingerprint

    Experimental design
    Design of experiments
    Polynomials
    Linear regression
    Polynomial time
    Exponential time
    Approximation
    Generalise

    ASJC Scopus subject areas

    • Theoretical Computer Science
    • Computer Science(all)

    Cite this

    Horel, T., Ioannidis, S., & Muthukrishnan, S. (2014). Budget feasible mechanisms for experimental design. In LATIN 2014: Theoretical Informatics - 11th Latin American Symposium, Proceedings (pp. 719-730). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8392 LNCS). Springer-Verlag. https://doi.org/10.1007/978-3-642-54423-1_62

    Budget feasible mechanisms for experimental design. / Horel, Thibaut; Ioannidis, Stratis; Muthukrishnan, Shanmugavelayutham.

    LATIN 2014: Theoretical Informatics - 11th Latin American Symposium, Proceedings. Springer-Verlag, 2014. p. 719-730 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8392 LNCS).

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

    Horel, T, Ioannidis, S & Muthukrishnan, S 2014, Budget feasible mechanisms for experimental design. in LATIN 2014: Theoretical Informatics - 11th Latin American Symposium, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 8392 LNCS, Springer-Verlag, pp. 719-730, 11th Latin American Theoretical Informatics Symposium, LATIN 2014, Montevideo, Uruguay, 3/31/14. https://doi.org/10.1007/978-3-642-54423-1_62
    Horel T, Ioannidis S, Muthukrishnan S. Budget feasible mechanisms for experimental design. In LATIN 2014: Theoretical Informatics - 11th Latin American Symposium, Proceedings. Springer-Verlag. 2014. p. 719-730. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-54423-1_62
    Horel, Thibaut ; Ioannidis, Stratis ; Muthukrishnan, Shanmugavelayutham. / Budget feasible mechanisms for experimental design. LATIN 2014: Theoretical Informatics - 11th Latin American Symposium, Proceedings. Springer-Verlag, 2014. pp. 719-730 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
    @inproceedings{83c8415735fa4800b7e8eb4992b3a3ed,
    title = "Budget feasible mechanisms for experimental design",
    abstract = "We present a deterministic, polynomial time, budget feasible mechanism scheme, that is approximately truthful and yields a constant (≈ 12.98) factor approximation for the Experimental Design Problem (EDP). By applying previous work on budget feasible mechanisms with a submodular objective, one could only have derived either an exponential time deterministic mechanism or a randomized polynomial time mechanism. We also establish that no truthful, budget-feasible mechanism is possible within a factor 2 approximation, and show how to generalize our approach to a wide class of learning problems, beyond linear regression.",
    author = "Thibaut Horel and Stratis Ioannidis and Shanmugavelayutham Muthukrishnan",
    year = "2014",
    month = "1",
    day = "1",
    doi = "10.1007/978-3-642-54423-1_62",
    language = "English (US)",
    isbn = "9783642544224",
    series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
    publisher = "Springer-Verlag",
    pages = "719--730",
    booktitle = "LATIN 2014",

    }

    TY - GEN

    T1 - Budget feasible mechanisms for experimental design

    AU - Horel, Thibaut

    AU - Ioannidis, Stratis

    AU - Muthukrishnan, Shanmugavelayutham

    PY - 2014/1/1

    Y1 - 2014/1/1

    N2 - We present a deterministic, polynomial time, budget feasible mechanism scheme, that is approximately truthful and yields a constant (≈ 12.98) factor approximation for the Experimental Design Problem (EDP). By applying previous work on budget feasible mechanisms with a submodular objective, one could only have derived either an exponential time deterministic mechanism or a randomized polynomial time mechanism. We also establish that no truthful, budget-feasible mechanism is possible within a factor 2 approximation, and show how to generalize our approach to a wide class of learning problems, beyond linear regression.

    AB - We present a deterministic, polynomial time, budget feasible mechanism scheme, that is approximately truthful and yields a constant (≈ 12.98) factor approximation for the Experimental Design Problem (EDP). By applying previous work on budget feasible mechanisms with a submodular objective, one could only have derived either an exponential time deterministic mechanism or a randomized polynomial time mechanism. We also establish that no truthful, budget-feasible mechanism is possible within a factor 2 approximation, and show how to generalize our approach to a wide class of learning problems, beyond linear regression.

    UR - http://www.scopus.com/inward/record.url?scp=84899921103&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84899921103&partnerID=8YFLogxK

    U2 - 10.1007/978-3-642-54423-1_62

    DO - 10.1007/978-3-642-54423-1_62

    M3 - Conference contribution

    AN - SCOPUS:84899921103

    SN - 9783642544224

    T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

    SP - 719

    EP - 730

    BT - LATIN 2014

    PB - Springer-Verlag

    ER -