NIPS 1998 Workshop

Large Margin Classifiers


Many pattern classifiers are represented as thresholded real-valued functions, eg: sigmoid neural networks, support vector machines, voting classifiers, and Bayesian schemes. There is currently a great deal of interest in algorithms that produce classifiers of this kind with large margins, where the margin is the amount by which the classifier's prediction is to the correct side of threshold. Recent theoretical and experimental results show that many learning algorithms (such as back-propagation, SVM methods, AdaBoost, and bagging) frequently produce classifiers with large margins, and that this leads to better generalization performance. Hence there is sufficient reason to believe that Large Margin Classifiers will become a core method of the standard machine learning toolbox.

The aims of this workshop are:

  • To provide an overview of recent developments in large margin classifiers, ranging from theoretical results to applications.

  • To explore connections with other methods (e.g. Bayesian inference).

  • To identify weaknesses, strengths and directions for future research in the area of large margin classifiers.


  • Mathematical Programming

  • Implicit methods for coping with large dimensional problems (Support Vectors and Kernels)

  • Iterative Ensemble Methods (Boosting, Bagging, Arcing, etc.)

  • Organized Panel Discussion with major researchers in the fields mentioned above. Prospective list of participants

    • Leo Breiman, Stanford University

    • Yoav Freund, AT&T Research

    • Olvi Mangasarian, University of Madison

    • V. Vapnik, AT&T Research

    • Manfred Warmuth, UC Santa Cruz

Special interest will be paid to the following issues

  • Theory of generalization and regularization

  • Algorithms

  • Applications

  • Benchmarks

Date and Location

Due to the large interest the workshop will be held on two days, namely Friday and Saturday, December 4-5 1998 in Breckenridge, CO.

Getting in Touch

If you are interested in contributing please fill out the survey form, send e-mail to [email protected] or fax to +49-30-6392-1805 (Alex Smola)

Click here for the guidelines for authors regarding formatting and style of the contributions.