Distributed Saddle-Point Problems Under Similarity

arXiv2107.10706MaRDI QIDQ6373497

Author name not available (Why is that?)

Publication date: 22 July 2021

Abstract: We study solution methods for (strongly-)convex-(strongly)-concave Saddle-Point Problems (SPPs) over networks of two type - master/workers (thus centralized) architectures and meshed (thus decentralized) networks. The local functions at each node are assumed to be similar, due to statistical data similarity or otherwise. We establish lower complexity bounds for a fairly general class of algorithms solving the SPP. We show that a given suboptimality $epsilon>0$ is achieved over master/workers networks in $$ rounds of communications, where $delta>0$ measures the degree of similarity of the local functions, $mu$ is their strong convexity constant, and $Delta$ is the diameter of the network. The lower communication complexity bound over meshed networks reads $$, where $ho$ is the (normalized) eigengap of the gossip matrix used for the communication between neighbouring nodes. We then propose algorithms matching the lower bounds over either types of networks (up to log-factors). We assess the effectiveness of the proposed algorithms on a robust logistic regression problem.

Has companion code repository: https://github.com/alexrogozin12/data_sim_sp

This page was built for publication: Distributed Saddle-Point Problems Under Similarity

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6373497)