Bolte, Jérôme and Pauwels, Edouard (2019) Conservative set valued fields, automatic differentiation, stochastic gradient methods and deep learning. Mathematical Programming. pp. 1-33. (In Press)
Preview |
Text
Download (486kB) | Preview |
Abstract
Modern problems in AI or in numerical analysis require nonsmooth approaches with a
exible calculus. We introduce generalized derivatives called conservative
fields for which we develop a calculus and provide representation formulas. Functions having a conservative field are called path differentiable: convex, concave,
Clarke regular and any semialgebraic Lipschitz continuous functions are path differentiable. Using Whitney stratification techniques for semialgebraic and definable
sets, our model provides variational formulas for nonsmooth automatic diffrentiation oracles, as for instance the famous backpropagation algorithm in deep learning.
Our differential model is applied to establish the convergence in values of nonsmooth stochastic gradient methods as they are implemented in practice.
Item Type: | Article |
---|---|
Language: | English |
Date: | 7 October 2019 |
Refereed: | Yes |
Place of Publication: | Heidelberg |
Uncontrolled Keywords: | Deep Learning, Automatic differentiation, Backpropagation algorithm, Nonsmooth stochastic optimization, Defiable sets, o-minimal structures, Stochastic gradient, Clarke subdifferential, First order methods |
Subjects: | B- ECONOMIE ET FINANCE |
Divisions: | TSE-R (Toulouse) |
Site: | UT1 |
Date Deposited: | 25 Mar 2021 10:11 |
Last Modified: | 23 Jul 2021 13:06 |
OAI Identifier: | oai:tse-fr.eu:125180 |
URI: | https://publications.ut-capitole.fr/id/eprint/42324 |