Stochastic Composite Least-Squares Regression with Convergence Rate O(1/n)

Nicolas Flammarion 1, 2 Francis Bach 1, 2
2 SIERRA - Statistical Machine Learning and Parsimony
DI-ENS - Département d'informatique de l'École normale supérieure, CNRS - Centre National de la Recherche Scientifique, Inria de Paris
Abstract : We consider the optimization of a quadratic objective function whose gradients are only accessible through a stochastic oracle that returns the gradient at any given point plus a zero-mean finite variance random error. We present the first algorithm that achieves jointly the optimal prediction error rates for least-squares regression, both in terms of forgetting of initial conditions in O(1/n 2), and in terms of dependence on the noise and dimension d of the problem, as O(d/n). Our new algorithm is based on averaged accelerated regularized gradient descent, and may also be analyzed through finer assumptions on initial conditions and the Hessian matrix, leading to dimension-free quantities that may still be small while the " optimal " terms above are large. In order to characterize the tightness of these new bounds, we consider an application to non-parametric regression and use the known lower bounds on the statistical performance (without computational limits), which happen to match our bounds obtained from a single pass on the data and thus show optimality of our algorithm in a wide variety of particular trade-offs between bias and variance.
Type de document :
Communication dans un congrès
Proceedings of The 30th Conference on Learning Theory, (COLT), 2017, Amsterdam, Netherlands. 2017
Liste complète des métadonnées

Littérature citée [60 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-01472867
Contributeur : Nicolas Flammarion <>
Soumis le : mardi 21 février 2017 - 13:57:32
Dernière modification le : jeudi 26 avril 2018 - 10:29:13
Document(s) archivé(s) le : lundi 22 mai 2017 - 15:24:03

Fichiers

hal-mdh.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01472867, version 1
  • ARXIV : 1602.05419

Collections

Citation

Nicolas Flammarion, Francis Bach. Stochastic Composite Least-Squares Regression with Convergence Rate O(1/n). Proceedings of The 30th Conference on Learning Theory, (COLT), 2017, Amsterdam, Netherlands. 2017. 〈hal-01472867〉

Partager

Métriques

Consultations de la notice

520

Téléchargements de fichiers

773