English

Implicit Rugosity Regularization via Data Augmentation

Machine Learning 2019-10-14 v3 Machine Learning

Abstract

Deep (neural) networks have been applied productively in a wide range of supervised and unsupervised learning tasks. Unlike classical machine learning algorithms, deep networks typically operate in the \emph{overparameterized} regime, where the number of parameters is larger than the number of training data points. Consequently, understanding the generalization properties and the role of (explicit or implicit) regularization in these networks is of great importance. In this work, we explore how the oft-used heuristic of \emph{data augmentation} imposes an {\em implicit regularization} penalty of a novel measure of the \emph{rugosity} or "roughness" based on the tangent Hessian of the function fit to the training data.

Keywords

Cite

@article{arxiv.1905.11639,
  title  = {Implicit Rugosity Regularization via Data Augmentation},
  author = {Daniel LeJeune and Randall Balestriero and Hamid Javadi and Richard G. Baraniuk},
  journal= {arXiv preprint arXiv:1905.11639},
  year   = {2019}
}

Comments

15 pages, 12 figures