A theoretical and experimental account of n-tuple classifier performance

Richard Rohwer*, Michal Morciniec

*Corresponding author for this work

    Research output: Contribution to journalArticle

    Abstract

    The n-tuple recognition method is briefly reviewed, summarizing the main theoretical results. Large-scale experiments carried out on Stat-Log project datasets confirm this method as a viable competitor to more popular methods due to its speed, simplicity, and accuracy on the majority of a wide variety of classification problems. A further investigation into the failure of the method on certain datasets finds the problem to be largely due to a mismatch between the scales which describe generalization and data sparseness.

    Original languageEnglish
    Pages (from-to)629-642
    Number of pages14
    JournalNeural Computation
    Volume8
    Issue number3
    Publication statusPublished - Apr 1996

    Fingerprint

    Classifier
    Datasets
    Mismatch
    Simplicity
    Experiment

    Keywords

    • n-tuple recognition
    • StatLog

    Cite this

    Rohwer, Richard ; Morciniec, Michal. / A theoretical and experimental account of n-tuple classifier performance. In: Neural Computation. 1996 ; Vol. 8, No. 3. pp. 629-642.
    @article{b000514d217746f081b8f76f3a036bf6,
    title = "A theoretical and experimental account of n-tuple classifier performance",
    abstract = "The n-tuple recognition method is briefly reviewed, summarizing the main theoretical results. Large-scale experiments carried out on Stat-Log project datasets confirm this method as a viable competitor to more popular methods due to its speed, simplicity, and accuracy on the majority of a wide variety of classification problems. A further investigation into the failure of the method on certain datasets finds the problem to be largely due to a mismatch between the scales which describe generalization and data sparseness.",
    keywords = "n-tuple recognition, StatLog",
    author = "Richard Rohwer and Michal Morciniec",
    year = "1996",
    month = "4",
    language = "English",
    volume = "8",
    pages = "629--642",
    journal = "Neural Computation",
    issn = "0899-7667",
    publisher = "MIT Press Journals",
    number = "3",

    }

    Rohwer, R & Morciniec, M 1996, 'A theoretical and experimental account of n-tuple classifier performance', Neural Computation, vol. 8, no. 3, pp. 629-642.

    A theoretical and experimental account of n-tuple classifier performance. / Rohwer, Richard; Morciniec, Michal.

    In: Neural Computation, Vol. 8, No. 3, 04.1996, p. 629-642.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - A theoretical and experimental account of n-tuple classifier performance

    AU - Rohwer, Richard

    AU - Morciniec, Michal

    PY - 1996/4

    Y1 - 1996/4

    N2 - The n-tuple recognition method is briefly reviewed, summarizing the main theoretical results. Large-scale experiments carried out on Stat-Log project datasets confirm this method as a viable competitor to more popular methods due to its speed, simplicity, and accuracy on the majority of a wide variety of classification problems. A further investigation into the failure of the method on certain datasets finds the problem to be largely due to a mismatch between the scales which describe generalization and data sparseness.

    AB - The n-tuple recognition method is briefly reviewed, summarizing the main theoretical results. Large-scale experiments carried out on Stat-Log project datasets confirm this method as a viable competitor to more popular methods due to its speed, simplicity, and accuracy on the majority of a wide variety of classification problems. A further investigation into the failure of the method on certain datasets finds the problem to be largely due to a mismatch between the scales which describe generalization and data sparseness.

    KW - n-tuple recognition

    KW - StatLog

    UR - http://www.scopus.com/inward/record.url?scp=0005646696&partnerID=8YFLogxK

    UR - http://www.mitpressjournals.org/doi/abs/10.1162/neco.1996.8.3.629

    M3 - Article

    VL - 8

    SP - 629

    EP - 642

    JO - Neural Computation

    JF - Neural Computation

    SN - 0899-7667

    IS - 3

    ER -