Percentile objective criteria in limiting average Markov control problems

Jerzy A. Filar, Dmitry Krass, Keith Ross

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

    Abstract

    The point of view is adopted that there are many natural situations in which the controller is interested in finding a policy that will achieve a sufficiently high long-run average reward, that is, a target level with a sufficiently high probability, that is, a percentile. The key conceptual difference between this approach and the classical problem is that the present controller is not searching for an optimal policy but rather for a policy that is 'good enough,' knowing that such a policy will typically fail to exist if the target level and the percentile are set too high. Conceptually, the approach is somewhat analogous to that often adopted by statisticians in testing of hypotheses where it is desirable (but usually not possible) to minimize simultaneously the so-called 'type 1' and the 'type 2' errors.

    Original languageEnglish (US)
    Title of host publicationProceedings of the IEEE Conference on Decision and Control
    PublisherPubl by IEEE
    Pages1273-1276
    Number of pages4
    Volume2
    StatePublished - 1989
    EventProceedings of the 28th IEEE Conference on Decision and Control. Part 2 (of 3) - Tampa, FL, USA
    Duration: Dec 13 1989Dec 15 1989

    Other

    OtherProceedings of the 28th IEEE Conference on Decision and Control. Part 2 (of 3)
    CityTampa, FL, USA
    Period12/13/8912/15/89

    Fingerprint

    Controllers
    Testing

    ASJC Scopus subject areas

    • Chemical Health and Safety
    • Control and Systems Engineering
    • Safety, Risk, Reliability and Quality

    Cite this

    Filar, J. A., Krass, D., & Ross, K. (1989). Percentile objective criteria in limiting average Markov control problems. In Proceedings of the IEEE Conference on Decision and Control (Vol. 2, pp. 1273-1276). Publ by IEEE.

    Percentile objective criteria in limiting average Markov control problems. / Filar, Jerzy A.; Krass, Dmitry; Ross, Keith.

    Proceedings of the IEEE Conference on Decision and Control. Vol. 2 Publ by IEEE, 1989. p. 1273-1276.

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

    Filar, JA, Krass, D & Ross, K 1989, Percentile objective criteria in limiting average Markov control problems. in Proceedings of the IEEE Conference on Decision and Control. vol. 2, Publ by IEEE, pp. 1273-1276, Proceedings of the 28th IEEE Conference on Decision and Control. Part 2 (of 3), Tampa, FL, USA, 12/13/89.
    Filar JA, Krass D, Ross K. Percentile objective criteria in limiting average Markov control problems. In Proceedings of the IEEE Conference on Decision and Control. Vol. 2. Publ by IEEE. 1989. p. 1273-1276
    Filar, Jerzy A. ; Krass, Dmitry ; Ross, Keith. / Percentile objective criteria in limiting average Markov control problems. Proceedings of the IEEE Conference on Decision and Control. Vol. 2 Publ by IEEE, 1989. pp. 1273-1276
    @inproceedings{1e1f3c9b99bc48b0bd33a237bc083d0e,
    title = "Percentile objective criteria in limiting average Markov control problems",
    abstract = "The point of view is adopted that there are many natural situations in which the controller is interested in finding a policy that will achieve a sufficiently high long-run average reward, that is, a target level with a sufficiently high probability, that is, a percentile. The key conceptual difference between this approach and the classical problem is that the present controller is not searching for an optimal policy but rather for a policy that is 'good enough,' knowing that such a policy will typically fail to exist if the target level and the percentile are set too high. Conceptually, the approach is somewhat analogous to that often adopted by statisticians in testing of hypotheses where it is desirable (but usually not possible) to minimize simultaneously the so-called 'type 1' and the 'type 2' errors.",
    author = "Filar, {Jerzy A.} and Dmitry Krass and Keith Ross",
    year = "1989",
    language = "English (US)",
    volume = "2",
    pages = "1273--1276",
    booktitle = "Proceedings of the IEEE Conference on Decision and Control",
    publisher = "Publ by IEEE",

    }

    TY - GEN

    T1 - Percentile objective criteria in limiting average Markov control problems

    AU - Filar, Jerzy A.

    AU - Krass, Dmitry

    AU - Ross, Keith

    PY - 1989

    Y1 - 1989

    N2 - The point of view is adopted that there are many natural situations in which the controller is interested in finding a policy that will achieve a sufficiently high long-run average reward, that is, a target level with a sufficiently high probability, that is, a percentile. The key conceptual difference between this approach and the classical problem is that the present controller is not searching for an optimal policy but rather for a policy that is 'good enough,' knowing that such a policy will typically fail to exist if the target level and the percentile are set too high. Conceptually, the approach is somewhat analogous to that often adopted by statisticians in testing of hypotheses where it is desirable (but usually not possible) to minimize simultaneously the so-called 'type 1' and the 'type 2' errors.

    AB - The point of view is adopted that there are many natural situations in which the controller is interested in finding a policy that will achieve a sufficiently high long-run average reward, that is, a target level with a sufficiently high probability, that is, a percentile. The key conceptual difference between this approach and the classical problem is that the present controller is not searching for an optimal policy but rather for a policy that is 'good enough,' knowing that such a policy will typically fail to exist if the target level and the percentile are set too high. Conceptually, the approach is somewhat analogous to that often adopted by statisticians in testing of hypotheses where it is desirable (but usually not possible) to minimize simultaneously the so-called 'type 1' and the 'type 2' errors.

    UR - http://www.scopus.com/inward/record.url?scp=0024915757&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=0024915757&partnerID=8YFLogxK

    M3 - Conference contribution

    VL - 2

    SP - 1273

    EP - 1276

    BT - Proceedings of the IEEE Conference on Decision and Control

    PB - Publ by IEEE

    ER -