• DocumentCode
    1465160
  • Title

    The one-inclusion graph algorithm is near-optimal for the prediction model of learning

  • Author

    Li, Yi ; Long, Philip M. ; Srinivasan, Aravind

  • Author_Institution
    Dept. of Comput. Sci., Nat. Univ. of Singapore, Singapore
  • Volume
    47
  • Issue
    3
  • fYear
    2001
  • fDate
    3/1/2001 12:00:00 AM
  • Firstpage
    1257
  • Lastpage
    1261
  • Abstract
    Haussler, Littlestone and Warmuth (1994) described a general-purpose algorithm for learning according to the prediction model, and proved an upper bound on the probability that their algorithm makes a mistake in terms of the number of examples seen and the Vapnik-Chervonenkis (VC) dimension of the concept class being learned. We show that their bound is within a factor of 1+o(1) of the best possible such bound for any algorithm
  • Keywords
    graph theory; learning systems; optimisation; prediction theory; probability; Vapnik-Chervonenkis dimension; concept class; general-purpose algorithm; learning prediction model; near-optimal algorithm; one-inclusion graph algorithm; probability; upper bound; Computer science; Neural networks; Prediction algorithms; Predictive models; Probability distribution; Upper bound; Virtual colonoscopy;
  • fLanguage
    English
  • Journal_Title
    Information Theory, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0018-9448
  • Type

    jour

  • DOI
    10.1109/18.915700
  • Filename
    915700