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

Download:

Current browse context:

math.OC

Change to browse by:

References & Citations

Bookmark

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

Mathematics > Optimization and Control

Title: Optimal convergence rates of totally asynchronous optimization

Abstract: Asynchronous optimization algorithms are at the core of modern machine learning and resource allocation systems. However, most convergence results consider bounded information delays and several important algorithms lack guarantees when they operate under total asynchrony. In this paper, we derive explicit convergence rates for the proximal incremental aggregated gradient (PIAG) and the asynchronous block-coordinate descent (Async-BCD) methods under a specific model of total asynchrony, and show that the derived rates are order-optimal. The convergence bounds provide an insightful understanding of how the growth rate of the delays deteriorates the convergence times of the algorithms. Our theoretical findings are demonstrated by a numerical example.
Comments: 6 pages
Subjects: Optimization and Control (math.OC)
Cite as: arXiv:2203.04611 [math.OC]
  (or arXiv:2203.04611v1 [math.OC] for this version)

Submission history

From: Xuyang Wu [view email]
[v1] Wed, 9 Mar 2022 09:58:48 GMT (200kb)

Link back to: arXiv, form interface, contact.