Abstract
It is known that the covering numbers of a function class on a double sample (length 2m) can be used to bound the generalization performance of a classifier by using a margin based analysis. In this paper we show that one can utilize an analogous argument in terms of the observed covering numbers on a single m-sample (being the actual observed data points). The significance of this is that for certain interesting classes of functions, such as support vector machines, there are new techniques which allow one to find good estimates for such covering numbers in terms of the speed of decay of the eigenvalues of a Gram matrix. These covering numbers can be much less than a priori bounds indicate in situations where the particular data received is “easy”. The work can be considered an extension of previous results which provided generalization performance bounds in terms of the VC-dimension of the class of hypotheses restricted to the sample, with the considerable advantage that the covering numbers can be readily computed, and they often are small.
This work was supported by the Australian Research Council and the European Commission under the Working Group Nr. 27150 (NeuroCOLT2).
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
N. Alon, S. Ben-David, N. Cesa-Bianchi and D. Haussler, “Scale-sensitive Dimensions, Uniform Convergence, and Learnability,” Journal of the ACM 44(4), 615–631, (1997).
Peter L. Bartlett, “The Sample Complexity of Pattern Classification with Neural Networks: The Size of the Weights is More Important than the Size of the Network,” IEEE Trans. Inf. Theory, 44(2), 525–536, (1998).
C. Cortes and V. Vapnik, “Support-Vector Networks,” Machine Learning, 20, 273–297 (1995).
Michael J. Kearns and Robert E. Schapire, “Efficient Distribution-free Learning of Probabilistic Concepts,” pages 382–391 in Proceedings of the 31st Symposium on the Foundations of Computer Science, IEEE Computer Society Press, Los Alamitos, CA, 1990.
Gábor Lugosi and Márta Pintér, “A Data-dependent Skeleton Estimate for Learning,” pages 51–56 in Proceedings of the Ninth Annual Workshop on Computational Learning Theory, Association for Computing Machinery, New York, 1996.
John Shawe-Taylor, Peter Bartlett, Robert Williamson and Martin Anthony, “A Framework for Structural Risk Minimization”, pages 68–76 in Proceedings of the 9th Annual Conference on Computational Learning Theory, Association for Computing Machinery, New York, 1996.
John Shawe-Taylor, Peter Bartlett, Robert Williamson and Martin Anthony, “Structural Risk Minimization over Data-Dependent Hierarchies”, IEEE Transactions on Information Theory, 44(5), 1926–1940 (1998).
Vladimir N. Vapnik, Estimation of Dependences Based on Empirical Data, Springer-Verlag, New York, 1982.
Vladimir N. Vapnik and Aleksei Ja. Chervonenkis, “On the Uniform Convergence of Relative Frequencies of Events to their Probabilities,” Theory of Probability and Applications, 16, 264–280 (1971).
Robert C. Williamson, Alex J. Smola and Bernhard Schölkopf, “Entropy Numbers, Operators and Support Vector Kernels,” EuroCOLT’99 (these proceedings). See also “Generalization Performance of Regularization Networks and Support Vector Machines via Entropy Numbers of Compact Operators,” http://spigot.anu.edu.au/~williams/papers/P100.ps submitted to IEEE Transactions on Information Theory, July 1998.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1999 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Shawe-Taylor, J., Williamson, R.C. (1999). Generalization Performance of Classifiers in Terms of Observed Covering Numbers. In: Fischer, P., Simon, H.U. (eds) Computational Learning Theory. EuroCOLT 1999. Lecture Notes in Computer Science(), vol 1572. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-49097-3_22
Download citation
DOI: https://doi.org/10.1007/3-540-49097-3_22
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-65701-9
Online ISBN: 978-3-540-49097-5
eBook Packages: Springer Book Archive