Breaking the Nonsmooth Barrier: A Scalable Parallel Method for Composite Optimization
Résumé
Due to their simplicity and excellent performance, parallel asynchronous variants of stochastic gradient descent have become popular methods to solve a wide range of large-scale optimization problems on multi-core architectures. Yet, despite their practical success, support for nonsmooth objectives is still lacking, making them unsuitable for many problems of interest in machine learning, such as the Lasso, group Lasso or empirical risk minimization with convex constraints. In this work, we propose and analyze ProxASAGA, a fully asynchronous sparse method inspired by SAGA, a variance reduced incremental gradient algorithm. The proposed method is easy to implement and significantly outperforms the state of the art on several nonsmooth, large-scale problems. We prove that our method achieves a theoretical linear speedup with respect to the sequential version under assumptions on the sparsity of gradients and block-separability of the proximal term. Empirical benchmarks on a multi-core architecture illustrate practical speedups of up to 12x on a 20-core machine.
Fichier principal
index.pdf (577.05 Ko)
Télécharger le fichier
index.bbl (8.83 Ko)
Télécharger le fichier
nips_2017.sty (10.11 Ko)
Télécharger le fichier
prox_asaga_1.pdf (44.65 Ko)
Télécharger le fichier
prox_asaga_2.pdf (27.07 Ko)
Télécharger le fichier
prox_asaga_dense_covtype.pdf (33.65 Ko)
Télécharger le fichier
prox_asaga_dense_rcv1.pdf (32.94 Ko)
Télécharger le fichier
prox_asaga_ideal_speedup.pdf (25.68 Ko)
Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Origine : Fichiers produits par l'(les) auteur(s)