Passer à la navigation principale Passer à la recherche Passer au contenu principal

Optimal algorithms for non-smooth distributed optimization in networks

  • Kevin Scaman
  • , Francis Bach
  • , Sébastien Bubeck
  • , Yin Tat Lee
  • , Laurent Massoulié
  • Huawei Noah's Ark Lab
  • Université PSL
  • Microsoft Research
  • University of Washington
  • MSR-Inria Joint Centre

Résultats de recherche: Contribution à un journalArticle de conférenceRevue par des pairs

121 Citations (Scopus)

Résumé

In this work, we consider the distributed optimization of non-smooth convex functions using a network of computing units. We investigate this problem under two regularity assumptions: (1) the Lipschitz continuity of the global objective function, and (2) the Lipschitz continuity of local individual functions. Under the local regularity assumption, we provide the first optimal first-order decentralized algorithm called multi-step primal-dual (MSPD) and its corresponding optimal convergence rate. A notable aspect of this result is that, for non-smooth functions, while the dominant term of the error is in O(1/t), the structure of the communication network only impacts a second-order term in O(1/t), where t is time. In other words, the error due to limits in communication resources decreases at a fast rate even in the case of non-strongly-convex objective functions. Under the global regularity assumption, we provide a simple yet efficient algorithm called distributed randomized smoothing (DRS) based on a local smoothing of the objective function, and show that DRS is within a d1/4 multiplicative factor of the optimal convergence rate, where d is the underlying dimension.

langue originaleAnglais
Pages (de - à)2740-2749
Nombre de pages10
journalAdvances in Neural Information Processing Systems
Volume2018-December
étatPublié - 1 janv. 2018
Modification externeOui
Evénement32nd Conference on Neural Information Processing Systems, NeurIPS 2018 - Montreal, Canada
Durée: 2 déc. 20188 déc. 2018

Empreinte digitale

Examiner les sujets de recherche de « Optimal algorithms for non-smooth distributed optimization in networks ». Ensemble, ils forment une empreinte digitale unique.

Contient cette citation