A nonlinear conjugate gradient method with complexity guarantees and its application to nonconvex regression
Chan-Renous-Legoubin, Rémi; Royer, Clément W. (2022), A nonlinear conjugate gradient method with complexity guarantees and its application to nonconvex regression, EURO Journal on Computational Optimization, 10, p. 100044. 10.1016/j.ejco.2022.100044
TypeArticle accepté pour publication ou publié
External document linkhttps://arxiv.org/pdf/2201.08568.pdf
Journal nameEURO Journal on Computational Optimization
MetadataShow full item record
Royer, Clément W.
Laboratoire d'analyse et modélisation de systèmes pour l'aide à la décision [LAMSADE]
Abstract (EN)Nonlinear conjugate gradients are among the most popular techniques for solving continuous optimization problems. Although these schemes have long been studied from a global convergence standpoint, their worst-case complexity properties have yet to be fully understood, especially in the nonconvex setting. In particular, it is unclear whether nonlinear conjugate gradient methods possess better guarantees than first-order methods such as gradient descent. Meanwhile, recent experiments have shown impressive performance of standard nonlinear conjugate gradient techniques on certain nonconvex problems, even when compared with methods endowed with the best known complexity guarantees.In this paper, we propose a nonlinear conjugate gradient scheme based on a simple line-search paradigm and a modified restart condition. These two ingredients allow for monitoring the properties of the search directions, which is instrumental in obtaining complexity guarantees. Our complexity results illustrate the possible discrepancy between nonlinear conjugate gradient methods and classical gradient descent. A numerical investigation on nonconvex robust regression problems as well as a standard benchmark illustrate that the restarting condition can track the behavior of a standard implementation.
Subjects / KeywordsOptimization and Control
Showing items related by title and author.
Curtis, F. E.; Robinson, D. P.; Royer, Clément; Wright, S. J. (2019) Article accepté pour publication ou publié
Bergou, E. H.; Diouane, Y.; Kungurtsev, V.; Royer, Clément W. (2022) Article accepté pour publication ou publié
A Liouville theorem for vector valued semilinear heat equations with no gradient structure and applications to blow-up Zaag, Hatem; Nouaili, Nejla (2010) Article accepté pour publication ou publié
Spectral Pollution and How to Avoid It (With Applications to Dirac and Periodic Schrödinger Operators) Lewin, Mathieu; Séré, Eric (2009) Article accepté pour publication ou publié
Bartier, Jean-Philippe; Laurençot, Philippe (2007) Article accepté pour publication ou publié