Header logo is ei

Local Rademacher Complexities

2005

Article

ei


We propose new bounds on the error of learning algorithms in terms of a data-dependent notion of complexity. The estimates we establish give optimal rates and are based on a local and empirical version of Rademacher averages, in the sense that the Rademacher averages are computed from the data, on a subset of functions with small empirical error. We present some applications to classification and prediction with convex function classes, and with kernel classes in particular.

Author(s): Bartlett, P. and Bousquet, O. and Mendelson, S.
Journal: The Annals of Statistics
Volume: 33
Number (issue): 4
Pages: 1497-1537
Year: 2005
Month: August
Day: 0

Department(s): Empirical Inference
Bibtex Type: Article (article)

Digital: 0
Language: en
Organization: Max-Planck-Gesellschaft
School: Biologische Kybernetik

Links: PDF
PostScript
Web

BibTex

@article{2000,
  title = {Local Rademacher Complexities},
  author = {Bartlett, P. and Bousquet, O. and Mendelson, S.},
  journal = {The Annals of Statistics},
  volume = {33},
  number = {4},
  pages = {1497-1537},
  organization = {Max-Planck-Gesellschaft},
  school = {Biologische Kybernetik},
  month = aug,
  year = {2005},
  month_numeric = {8}
}