We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Mathematics > Optimization and Control

Title: Dynamic Regret of Distributed Online Frank-Wolfe Convex Optimization

Abstract: This paper considers distributed online convex constrained optimization, in which various agents in a multi-agent system cooperate to minimize a global cost function through communicating with neighbors over a time-varying network. When the constraint set of optimization problem is high-dimensional and complicated, the computational cost of the projection operation often becomes prohibitive. To handle this problem, we develop a distributed online Frank-Wolfe optimization algorithm combining with gradient tracking technique. We rigorously establish the dynamic regret bound of the proposed optimization algorithm as $\mathcal{O}(\sqrt{T(1+H_T)}+D_T)$, which explicitly depends on the iteration round $T$, function variation $H_T$, and gradient variation $D_T$. Finally, the theoretical results are verified and compared in the case of distributed online ridge regression problems.
Subjects: Optimization and Control (math.OC)
Cite as: arXiv:2302.00663 [math.OC]
  (or arXiv:2302.00663v1 [math.OC] for this version)

Submission history

From: Wentao Zhang [view email]
[v1] Wed, 1 Feb 2023 18:48:45 GMT (163kb)

Link back to: arXiv, form interface, contact.