Improving the performance of radial basis function classifiers in condition monitoring and fault diagnosis applications where 'unknown' faults may occur

Yuhua Li, Michael J Pont

    Research output: Contribution to journal › Article

    32 Citations (Scopus)

    Abstract

    This paper develops a technique for determining a reliable threshold for RBF classifiers. A two-phase approach is proposed to RBF classifier use in situations where unknown faults may occur: the first phase deals with the possibility of unknown faults; in the second phase, the classifier threshold is modified through retraining using all available data, including newly collected data about unknown faults. The approach is easy to use and is demonstrated to be particularly effective in classification problems where novelty detection capability is required.
    LanguageEnglish
    Pages569-577
    JournalPattern Recognition Letter
    Volume23
    Issue number5
    DOIs
    Publication statusPublished - Mar 2002

    Fingerprint

    Condition monitoring
    Failure analysis
    Classifiers

    Cite this

    Li, Yuhua ; Pont, Michael J. / Improving the performance of radial basis function classifiers in condition monitoring and fault diagnosis applications where 'unknown' faults may occur. In: Pattern Recognition Letter. 2002 ; Vol. 23, No. 5. pp. 569-577.
    @article{d80f9836c3eb4a80b1e83a2b26a6c8bb,
    title = "Improving the performance of radial basis function classifiers in condition monitoring and fault diagnosis applications where 'unknown' faults may occur",
    abstract = "This paper develops a technique for determining a reliable threshold for RBF classifiers. A two-phase approach is proposed to RBF classifier use in situations where unknown faults may occur: the first phase deals with the possibility of unknown faults; in the second phase, the classifier threshold is modified through retraining using all available data, including newly collected data about unknown faults. The approach is easy to use and is demonstrated to be particularly effective in classification problems where novelty detection capability is required.",
    author = "Yuhua Li and Pont, {Michael J}",
    year = "2002",
    month = "3",
    doi = "10.1016/S0167-8655(01)00133-7",
    language = "English",
    volume = "23",
    pages = "569--577",
    journal = "Pattern Recognition Letters",
    issn = "0167-8655",
    publisher = "Elsevier",
    number = "5",

    }

    Li, Y & Pont, MJ 2002, 'Improving the performance of radial basis function classifiers in condition monitoring and fault diagnosis applications where 'unknown' faults may occur', Pattern Recognition Letter, vol. 23, no. 5, pp. 569-577. https://doi.org/10.1016/S0167-8655(01)00133-7

    Improving the performance of radial basis function classifiers in condition monitoring and fault diagnosis applications where 'unknown' faults may occur. / Li, Yuhua; Pont, Michael J.

    In: Pattern Recognition Letter, Vol. 23, No. 5, 03.2002, p. 569-577.

    Research output: Contribution to journal › Article

    TY - JOUR

    T1 - Improving the performance of radial basis function classifiers in condition monitoring and fault diagnosis applications where 'unknown' faults may occur

    AU - Li, Yuhua

    AU - Pont, Michael J

    PY - 2002/3

    Y1 - 2002/3

    N2 - This paper develops a technique for determining a reliable threshold for RBF classifiers. A two-phase approach is proposed to RBF classifier use in situations where unknown faults may occur: the first phase deals with the possibility of unknown faults; in the second phase, the classifier threshold is modified through retraining using all available data, including newly collected data about unknown faults. The approach is easy to use and is demonstrated to be particularly effective in classification problems where novelty detection capability is required.

    AB - This paper develops a technique for determining a reliable threshold for RBF classifiers. A two-phase approach is proposed to RBF classifier use in situations where unknown faults may occur: the first phase deals with the possibility of unknown faults; in the second phase, the classifier threshold is modified through retraining using all available data, including newly collected data about unknown faults. The approach is easy to use and is demonstrated to be particularly effective in classification problems where novelty detection capability is required.

    UR - http://www.elsevier.com/wps/find/journaldescription.cws_home/505619/description

    U2 - 10.1016/S0167-8655(01)00133-7

    DO - 10.1016/S0167-8655(01)00133-7

    M3 - Article

    VL - 23

    SP - 569

    EP - 577

    JO - Pattern Recognition Letters

    T2 - Pattern Recognition Letters

    JF - Pattern Recognition Letters

    SN - 0167-8655

    IS - 5

    ER -

    Li Y, Pont MJ. Improving the performance of radial basis function classifiers in condition monitoring and fault diagnosis applications where 'unknown' faults may occur. Pattern Recognition Letter. 2002 Mar;23(5):569-577. https://doi.org/10.1016/S0167-8655(01)00133-7

    Access to Document