Skip to main content

Stronger generalization bounds for deep nets via a compression approach

Author(s): Arora, Sanjeev; Ge, R; Neyshabur, B; Zhang, Y

To refer to this page use:
Abstract: Deep nets generalize well despite having more parameters than the number of training samples. Recent works try to give an explanation using PAC-Bayes and Margin-based analyses, but do not as yet result in sample complexity bounds better than naive parameter counting. The current paper shows generalization bounds that are orders of magnitude better in practice. These rely upon new succinct reparametrizations of the trained net - a compression that is explicit and efficient. These yield generalization bounds via a simple compression-based framework introduced here. Our results also provide some theoretical justification for widespread empirical success in compressing deep nets. Analysis of correctness of our compression relies upon some newly identified "noise stability"properties of trained deep nets, which are also experimentally verified. The study of these properties and resulting generalization bounds are also extended to convolutional nets, which had eluded earlier attempts on proving generalization.
Publication Date: 2018
Citation: Arora, S, Ge, R, Neyshabur, B, Zhang, Y. (2018). Stronger generalization bounds for deep nets via a compression approach. 1 (390 - 418
Pages: 390 - 418
Type of Material: Conference Article
Journal/Proceeding Title: 35th International Conference on Machine Learning, ICML 2018
Version: Author's manuscript

Items in OAR@Princeton are protected by copyright, with all rights reserved, unless otherwise indicated.