Selecting extreme patterns for classifier design

Yuhua Li

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

    Abstract

    Patterns in a data set have different levels of usefulness to the training of classifiers. Extreme patterns which are on the surface of the space occupied by the data are essential to the successful classifier design in some applications such as novelty detection. This paper presents a method that selects extreme patterns by analysing local geometrical information. It specifies a tangent haperplane at each pattern in order to determine the location of the pattern in the pattern space. A pattern is selected as an extreme pattern if it is identified to sit on the tangent hyperplane. The proposed method is evaluated using multilayer perceptrons and support vector machines on three benchmark classification problems. The experimental results demonstrate the proposed method is able to select extreme patterns for successful classifier design.
    LanguageEnglish
    Title of host publicationUnknown Host Publication
    Pages1259-1266
    Number of pages8
    Publication statusPublished - 11 Jun 2007
    EventThe World Congress on Engineering Asset Management and The International Conference on Condition Monitoring - Harrogate, UK
    Duration: 11 Jun 2007 → …

    Conference

    ConferenceThe World Congress on Engineering Asset Management and The International Conference on Condition Monitoring
    Period11/06/07 → …

    Fingerprint

    Classifiers
    Multilayer neural networks
    Support vector machines

    Cite this

    Li, Y. (2007). Selecting extreme patterns for classifier design. In Unknown Host Publication (pp. 1259-1266)
    Li, Yuhua. / Selecting extreme patterns for classifier design. Unknown Host Publication. 2007. pp. 1259-1266
    @inproceedings{79de8e8dc8064ab89740180c00c10632,
    title = "Selecting extreme patterns for classifier design",
    abstract = "Patterns in a data set have different levels of usefulness to the training of classifiers. Extreme patterns which are on the surface of the space occupied by the data are essential to the successful classifier design in some applications such as novelty detection. This paper presents a method that selects extreme patterns by analysing local geometrical information. It specifies a tangent haperplane at each pattern in order to determine the location of the pattern in the pattern space. A pattern is selected as an extreme pattern if it is identified to sit on the tangent hyperplane. The proposed method is evaluated using multilayer perceptrons and support vector machines on three benchmark classification problems. The experimental results demonstrate the proposed method is able to select extreme patterns for successful classifier design.",
    author = "Yuhua Li",
    year = "2007",
    month = "6",
    day = "11",
    language = "English",
    pages = "1259--1266",
    booktitle = "Unknown Host Publication",

    }

    Li, Y 2007, Selecting extreme patterns for classifier design. in Unknown Host Publication. pp. 1259-1266, The World Congress on Engineering Asset Management and The International Conference on Condition Monitoring, 11/06/07.

    Selecting extreme patterns for classifier design. / Li, Yuhua.

    Unknown Host Publication. 2007. p. 1259-1266.

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

    TY - GEN

    T1 - Selecting extreme patterns for classifier design

    AU - Li, Yuhua

    PY - 2007/6/11

    Y1 - 2007/6/11

    N2 - Patterns in a data set have different levels of usefulness to the training of classifiers. Extreme patterns which are on the surface of the space occupied by the data are essential to the successful classifier design in some applications such as novelty detection. This paper presents a method that selects extreme patterns by analysing local geometrical information. It specifies a tangent haperplane at each pattern in order to determine the location of the pattern in the pattern space. A pattern is selected as an extreme pattern if it is identified to sit on the tangent hyperplane. The proposed method is evaluated using multilayer perceptrons and support vector machines on three benchmark classification problems. The experimental results demonstrate the proposed method is able to select extreme patterns for successful classifier design.

    AB - Patterns in a data set have different levels of usefulness to the training of classifiers. Extreme patterns which are on the surface of the space occupied by the data are essential to the successful classifier design in some applications such as novelty detection. This paper presents a method that selects extreme patterns by analysing local geometrical information. It specifies a tangent haperplane at each pattern in order to determine the location of the pattern in the pattern space. A pattern is selected as an extreme pattern if it is identified to sit on the tangent hyperplane. The proposed method is evaluated using multilayer perceptrons and support vector machines on three benchmark classification problems. The experimental results demonstrate the proposed method is able to select extreme patterns for successful classifier design.

    M3 - Conference contribution

    SP - 1259

    EP - 1266

    BT - Unknown Host Publication

    ER -

    Li Y. Selecting extreme patterns for classifier design. In Unknown Host Publication. 2007. p. 1259-1266