ML Anthology

On Generalization Bounds, Projection Profile, and Margin Distribution

Ashutosh Garg, Sariel Har-Peled, Dan Roth
ICML 2002 pp. 171-178
/icml/2002/garg2002icml-generalization/

Abstract

We study generalization properties of linear learning algorithms and develop a data dependent approach that is used to derive generalization bounds that depend on the margin distribution. Our method makes use of random projection techniques to allow the use of existing VC dimension bounds in the effective, lower, dimension of the data. Comparisons with existing generalization bound show that our bounds are tighter and meaningful in cases existing bounds are not. 1

Semantic Scholar

Cite

Text

Garg et al. "On Generalization Bounds, Projection Profile, and Margin Distribution." International Conference on Machine Learning, 2002.

Markdown

[Garg et al. "On Generalization Bounds, Projection Profile, and Margin Distribution." International Conference on Machine Learning, 2002.](https://mlanthology.org/icml/2002/garg2002icml-generalization/)

BibTeX

@inproceedings{garg2002icml-generalization,
  title     = {{On Generalization Bounds, Projection Profile, and Margin Distribution}},
  author    = {Garg, Ashutosh and Har-Peled, Sariel and Roth, Dan},
  booktitle = {International Conference on Machine Learning},
  year      = {2002},
  pages     = {171-178},
  url       = {https://mlanthology.org/icml/2002/garg2002icml-generalization/}
}