Asynchronous sequential inertial iterations for common fixed points problems with an application to linear systems
The common fixed points problem requires finding a point in the intersection of fixed points sets of a finite collection of operators. Quickly solving problems of this sort is of great practical importance for engineering and scientific tasks (e.g., for computed tomography). Iterative methods for solving these problems often employ a Krasnosel’skiĭ–Mann type iteration. We present an asynchronous sequential inertial (ASI) algorithmic framework in a Hilbert space to solve common fixed points problems with a collection of nonexpansive operators. Our scheme allows use of out-of-date iterates when generating updates, thereby enabling processing nodes to work simultaneously and without synchronization. This method also includes inertial type extrapolation terms to increase the speed of convergence. In particular, we extend the application of the recent “ARock algorithm” (Peng et al. in SIAM J Sci Comput 38:A2851–A2879, 2016) in the context of convex feasibility problems. Convergence of the ASI algorithm is proven with no assumption on the distribution of delays, except that they be uniformly bounded. Discussion is provided along with a computational example showing the performance of the ASI algorithm applied in conjunction with a diagonally relaxed orthogonal projections (DROP) algorithm for estimating solutions to large linear systems.
KeywordsConvex feasibility problem Asynchronous sequential iterations Nonexpansive operator Fixed point iteration Kaczmarz method DROP algorithm
We thank Prof. Reinhard Schulte from Loma Linda Univeristy for valuable discussions and insights regarding the advantage of an asynchronous approach to large scale problems in proton CT (pCT) image reconstruction and Intensity-Modulated Proton Therapy (IMPT). We appreciate constructive comments received from Prof. Ernesto Gomez from California State University San Bernardino and Prof. Keith Schubert from Baylor University. We thank the anonymous reviewers for their constructive comments. We are indebted to Patrick Combettes, Jonathan Eckstein, Dan Gordon and Simeon Reich for valuable feedback on the earlier version of this work that was posted on arXiv. The first author’s work is supported by the National Science Foundation Graduate Research Fellowship under Grant No. DGE-1650604. Any opinion, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. The second author’s work was supported by research Grant No. 2013003 of the United States-Israel Binational Science Foundation (BSF).
- 7.Bauschke, H., Iorio, F., Koch, V.: The method of cyclic intrepid projections: convergence analysis and numerical experiments. In: Wakayama, M., Anderssen, R., Cheng, J., Fukumoto, Y., McKibbin, R., Polthier, K., Takagi, T., Toh, K.C. (eds.) The Impact of Applications on Mathematics, pp. 187–200. Springer Japan, Tokyo (2014)Google Scholar
- 15.Censor, Y., Elfving, T., Herman, G.: Averaging strings of sequential iterations for convex feasibility problems. In: Butnariu, D., Censor, Y., Reich, S. (eds.) Inherently Parallel Algorithms in Feasibility and Optimization and Their Applications, pp. 101–113. Elsevier, Amsterdam (2001)CrossRefGoogle Scholar
- 24.Combettes, P.: The Convex Feasibility Problem in Image Recovery, vol. 95. Academic Press, New York (1996)Google Scholar
- 27.Dong, Q.L., Huang, J.Z., Li, X.H., Cho, Y.J., Rassias, T.M.: MiKM: multi-step inertial Krasnoselski–Mann algorithm and its applications. J. Global Optim. (2018). https://doi.org/10.1007/s10898-018-0727-x
- 38.Johnstone, P.R., Eckstein, J.: Convergence Rates for Projective Splitting. arXiv preprint arXiv:1806.03920 (2018)
- 39.Johnstone, P.R., Eckstein, J.: Projective Splitting with Forward Steps: Asynchronous and Block-Iterative Operator Splitting. arXiv preprint arXiv:1803.07043 (2018)