[UAI] New paper in the Journal of Machine Learning Research: Bayes Point Machines

From: JMLR (cohn+jmlr@cs.cmu.edu)
Date: Tue Aug 21 2001 - 09:07:25 PDT

  • Next message: Ning Zhong: "[UAI] IEEE Data Mining 2001: Call for Participation"

    The Journal of Machine Learning Research (www.jmlr.org) is pleased to
    announce the availability of a new paper in electronic form.

    - ----------------------------------------
    Bayes Point Machines

    Ralf Herbrich, Thore Graepel and Colin Campbell. Journal of Machine Learning
    Research 1 (August 2001), pp. 245-279.

    Abstract

    Kernel-classifiers comprise a powerful class of non-linear decision
    functions for binary classification. The support vector machine is an
    example of a learning algorithm for kernel classifiers that singles out the
    consistent classifier with the largest margin, i.e. minimal real-valued
    output on the training sample, within the set of consistent hypotheses, the
    so-called version space. We suggest the Bayes point machine as a
    well-founded improvement which approximates the Bayes-optimal decision by
    the centre of mass of version space. We present two algorithms to
    stochastically approximate the centre of mass of version space: a billiard
    sampling algorithm and a sampling algorithm based on the well known
    perceptron algorithm. It is shown how both algorithms can be extended to
    allow for soft-boundaries in order to admit training errors. Experimentally,
    we find that - for the zero training error case - Bayes point machines
    consistently outperform support vector machines on both surrogate data and
    real-world benchmark data sets. In the soft-boundary/soft-margin case, the
    improvement over support vector machines is shown to be reduced. Finally, we
    demonstrate that the real-valued output of single Bayes points on novel test
    points is a valid confidence measure and leads to a steady decrease in
    generalisation error when used as a rejection criterion.

    This paper and earlier papers in Volume 1 are available electronically at
    http://www.jmlr.org in PostScript, PDF and HTML formats; a bound, hardcopy
    edition of Volume 1 will be available later this year.

    - -David Cohn, <david.cohn@acm.org>
      Managing Editor, Journal of Machine Learning Research

    -------
    This message has been sent to the mailing list "jmlr-announce@ai.mit.edu",
    which is maintained automatically by majordomo. To subscribe to the list,
    send mail to listserv@ai.mit.edu with the line "subscribe jmlr-announce" in
    the body; to unsubscribe send email to listserv@ai.mit.edu with the line
    "unsubscribe jmlr-announce" in the body.



    This archive was generated by hypermail 2b29 : Tue Aug 21 2001 - 09:14:06 PDT