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

Download:

Current browse context:

math.NA

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 > Numerical Analysis

Title: Additive Schwarz Methods for Convex Optimization with Backtracking

Authors: Jongho Park
Abstract: This paper presents a novel backtracking strategy for additive Schwarz methods for general convex optimization problems as an acceleration scheme. The proposed backtracking strategy is independent of local solvers, so that it can be applied to any algorithms that can be represented in an abstract framework of additive Schwarz methods. Allowing for adaptive increasing and decreasing of the step size along the iterations, the convergence rate of an algorithm is greatly improved. Improved convergence rate of the algorithm is proven rigorously. In addition, combining the proposed backtracking strategy with a momentum acceleration technique, we propose a further accelerated additive Schwarz method. Numerical results for various convex optimization problems that support our theory are presented.
Comments: 20 pages, 3 figures
Subjects: Numerical Analysis (math.NA); Optimization and Control (math.OC)
MSC classes: 65N55, 65B99, 65K15, 90C25
Cite as: arXiv:2110.07172 [math.NA]
  (or arXiv:2110.07172v1 [math.NA] for this version)

Submission history

From: Jongho Park [view email]
[v1] Thu, 14 Oct 2021 06:17:45 GMT (302kb,D)

Link back to: arXiv, form interface, contact.