Local Rademacher Complexity Bounds based on Covering Numbers
Artificial Intelligence
2015-10-07 v1 Machine Learning
Machine Learning
Abstract
This paper provides a general result on controlling local Rademacher complexities, which captures in an elegant form to relate the complexities with constraint on the expected norm to the corresponding ones with constraint on the empirical norm. This result is convenient to apply in real applications and could yield refined local Rademacher complexity bounds for function classes satisfying general entropy conditions. We demonstrate the power of our complexity bounds by applying them to derive effective generalization error bounds.
Cite
@article{arxiv.1510.01463,
title = {Local Rademacher Complexity Bounds based on Covering Numbers},
author = {Yunwen Lei and Lixin Ding and Yingzhou Bi},
journal= {arXiv preprint arXiv:1510.01463},
year = {2015}
}