English

Distributed Linearized Alternating Direction Method of Multipliers for Composite Convex Consensus Optimization

Optimization and Control 2017-01-03 v4

Abstract

Given an undirected graph G=(N,E)\mathcal{G}=(\mathcal{N},\mathcal{E}) of agents N={1,,N}\mathcal{N}=\{1,\ldots,N\} connected with edges in E\mathcal{E}, we study how to compute an optimal decision on which there is consensus among agents and that minimizes the sum of agent-specific private convex composite functions {Φi}iN\{\Phi_i\}_{i\in\mathcal{N}} while respecting privacy requirements, where Φiξi+fi\Phi_i\triangleq \xi_i + f_i belongs to agent-ii. Assuming only agents connected by an edge can communicate, we propose a distributed proximal gradient method DPGA for consensus optimization over both unweighted and weighted static (undirected) communication networks. In one iteration, each agent-ii computes the prox map of ξi\xi_i and gradient of fif_i, and this is followed by local communication with neighboring agents. We also study its stochastic gradient variant, SDPGA, which can only access to noisy estimates of fi\nabla f_i at each agent-ii. This computational model abstracts a number of applications in distributed sensing, machine learning and statistical inference. We show ergodic convergence in both sub-optimality error and consensus violation for DPGA and SDPGA with rates O(1/t)\mathcal{O}(1/t) and O(1/t)\mathcal{O}(1/\sqrt{t}), respectively.

Keywords

Cite

@article{arxiv.1512.08122,
  title  = {Distributed Linearized Alternating Direction Method of Multipliers for Composite Convex Consensus Optimization},
  author = {Necdet Serhat Aybat and Zi Wang and Tianyi Lin and Shiqian Ma},
  journal= {arXiv preprint arXiv:1512.08122},
  year   = {2017}
}

Comments

Theoretical results are improved and new numerical results are added

R2 v1 2026-06-22T12:18:16.933Z