Header logo is

Algorithmic Stability and Generalization Performance

2001

Conference Paper

ei


We present a novel way of obtaining PAC-style bounds on the generalization error of learning algorithms, explicitly using their stability properties. A {\em stable} learner being one for which the learned solution does not change much for small changes in the training set. The bounds we obtain do not depend on any measure of the complexity of the hypothesis space (e.g. VC dimension) but rather depend on how the learning algorithm searches this space, and can thus be applied even when the VC dimension in infinite. We demonstrate that regularization networks possess the required stability property and apply our method to obtain new bounds on their generalization performance.

Author(s): Bousquet, O. and Elisseeff, A.
Book Title: Advances in Neural Information Processing Systems 13
Journal: Advances in Neural Information Processing Systems
Pages: 196-202
Year: 2001
Month: April
Day: 0
Editors: Leen, T.K. , T.G. Dietterich, V. Tresp
Publisher: MIT Press

Department(s): Empirical Inference
Bibtex Type: Conference Paper (inproceedings)

Event Name: Fourteenth Annual Neural Information Processing Systems Conference (NIPS 2000)
Event Place: Denver, CO, USA

Address: Cambridge, MA, USA
Digital: 0
ISBN: 0-262-12241-3
Organization: Max-Planck-Gesellschaft
School: Biologische Kybernetik

Links: PDF
Web

BibTex

@inproceedings{1437,
  title = {Algorithmic Stability and Generalization Performance},
  author = {Bousquet, O. and Elisseeff, A.},
  journal = {Advances in Neural Information Processing Systems},
  booktitle = {Advances in Neural Information Processing Systems 13},
  pages = {196-202},
  editors = {Leen, T.K. , T.G. Dietterich, V. Tresp},
  publisher = {MIT Press},
  organization = {Max-Planck-Gesellschaft},
  school = {Biologische Kybernetik},
  address = {Cambridge, MA, USA},
  month = apr,
  year = {2001},
  month_numeric = {4}
}