Abstract
The alternating direction method of multipliers (ADMM) is a popular method for solving separable convex programs with linear constraints, and its proximal version is an important variant. In the literature, Fortin and Glowinski proved that the step size for updating the Lagrange multiplier of the ADMM can be chosen in the open interval of zero to the golden ratio, and subsequently this result has been proved to be also valid for the proximal ADMM. In this paper, we demonstrate that the dual step size can be larger than the golden ratio when the proximal regularization is positive definite. Thus, the feasible interval of the dual step size can be further enlarged for the proximal ADMM. Moreover, we establish the exact relationship between the dual step size and the proximal parameter. We also prove global convergence and establish a worst case convergence rate in the ergodic sense for this proximal scheme with the enlarged step size. Finally, we present numerical results to demonstrate the practical performance of the method.
Similar content being viewed by others
References
Hestenes, M.R.: Multiplier and gradient methods. J. Optim. Theory Appl. 4, 303–320 (1969)
Powell, M.J.D.: A method for nonlinear constraints in minimization problems. In: Fletcher, R (ed.) Optimization, pp 283–298. Academic Press, New York (1969)
Chan, T.F., Glowinski, R.: Finite element approximation and iterative solution of a class of mildly nonlinear elliptic equations. Stanford report STAN-CS-78-674, Computer Science Department, Stanford University, Palo Alto CA (1978)
Glowinski, R., Marroco, A.: Sur l’approximation, par éléments finis d’ordre un, et la résolution, par pénalisation-dualité d’une classe de problèmes de Dirichlet non linéaires. Revue franċaise d’automatique, informatique, recherche opérationnelle. Analyse numérique 9(2), 41–76 (1975)
Boyd, S., Parikh, N., Chu, E., Peleato, B., Eckstein, J.: Distributed optimization and statistical learning via the alternating direction method of multipliers. Found. Trends Mach. Learn. 3(1), 1–122 (2011)
Eckstein, J., Yao, W.: Understanding the convergence of the alternating direction method of multipliers: Theoretical and computational perspectives. Pac. J. Optim. 11(4), 619–644 (2015)
Glowinski, R.: On alternating directionmethods of multipliers: a historical perspective. In: Fitzgibbon, W., Kuznetsov, Y.A., Neittaanmaki, P., Pironneau, O. (eds.) Modeling, Simulation and Optimization for Science and Technology. Computational Methods in Applied Sciences, vol. 34, pp 59–82. Springer, Dordrecht (2014)
Glowinski, R., Pan, T.W., Tai, X.C.: Some facts about operator-splitting and alternating direction methods. In: Glowinski, R., Osher, S.J., Yin, W. (eds.) Splitting Methods in Communication, Imaging, Science, and Engineering, pp 19–94. Springer, New York (2016)
Glowinski, R., Osher, S.J., Yin, W. (eds.): Splitting Methods in Communication, Imaging, Science, and Engineering. Springer, New York (2016)
Cai, X.J., Gu, G.Y., He, B.S., Yuan, X.: A proximal point algorithm revisit on the alternating direction method of multipliers. Sci. China Math. 56(10), 2179–2186 (2013)
Ma, S.Q.: Alternating proximal gradient method for convex minimization. J. Sci. Comput. 68, 546–572 (2016)
Eckstein, J., Bertsekas, D.P.: On the Douglas–Rachford splitting method and the proximal point algorithm for maximal monotone operators. Math. Program. 55 (1-3), 293–318 (1992)
He, B.S., Liu, H., Wang, Z.R., Yuan, X.: A strictly contractive Peaceman–Rachford splitting method for convex programming. SIAM J. Optim. 24 (3), 1011–1040 (2014)
Ma, F.: On relaxation of some customized proximal point algorithms for convex minimization: from variational inequality perspective. Comput. Optim. Appl. 73(3), 871–901 (2019)
He, B., Ma, F., Yuan, X.: Convergence study on the symmetric version of ADMM with larger step sizes. SIAM J. Imaging Sci. 9(3), 1467–1501 (2016)
Fortin, M., Glowinski, R.: Méthodes de Lagrangien Augmenté: Application à la Résolution Numérique de Problèmes aux Limites, Dunod, Paris (1982)
Fortin, M., Glowinski, R.: Augmented Lagrangian Methods: Applications to the Numerical Solution of Boundary-Value Problems, Amsterdam the Netherlands (1983)
Glowinski, R.: Numerical Methods for Nonlinear Variational Problems. Springer, New York (1984)
Wen, Z.W., Goldfarb, D., Yin, W.T.: Alternating direction augmented Lagrangian methods for semidefinite programming. Math. Program. Comput. 2, 203–230 (2010)
Chen, C.H., He, B.S., Yuan, X.: Matrix completion via alternating direction method. IMA J. Numer. Analy. 32, 227–245 (2012)
He, B.S., Xu, M.H., Yuan, X.: Solving large-scale least squares covariance matrix problems by alternating direction methods. SIAM J. Matrix Anal. Appli. 32, 136–152 (2011)
Gabay, D., Mercier, B.: A dual algorithm for the solution of nonlinear variational problems via finite-element approximations. Comput. Math. Appli. 2, 17–40 (1976)
Tao, M., Yuan, X.: On Glowinski’s open question on the alternating direction method of multipliers. J. Optim. Theory Appl. 179, 163–196 (2018)
Eckstein, J.: Some saddle-function splitting methods for convex programming. Optim. Methods Soft. 4, 75–83 (1994)
He, B.S., Liao, L.Z., Han, D.R., Yang, H.: A new inexact alternating directions method for monotone variational inequalities. Math. Program. 92, 103–118 (2002)
Yang, J., Yuan, X.: Linearized augmented Lagrangian and alternating direction methods for nuclear norm minimization. Math. Comput. 82(281), 301–329 (2013)
Deng, W., Yin, W.T.: On the global and linear convergence of the generalized alternating direction method of multipliers. J. Sci. Comput. 66, 889–916 (2016)
Zhang, X.Q., Burger, M., Osher, S.: A unified primal-dual algorithm framework based on Bregman iteration. J. Sci. Comput. 96, 20–46 (2011)
Chan, R.H., Tao, M., Yuan, X.: Linearized alternating direction method of multipliers for constrained linear least-squares problem. East Asian J. Appl. Math. 2, 326–341 (2012)
Wang, X., Yuan, X.: The linearized alternating direction method of multipliers for dantzig selector. SIAM J. Sci. Comput. 34(5), A2792–A2811 (2012)
Xu, M.H.: Proximal alternating directions method for structured variational inequalities. J. Optim. Theory Appl. 134, 107–117 (2007)
Gonçalves, M. L. N., Melo, J.G., Monteiro, R.D.C.: Extending the ergodic convergence rate of the proximal ADMM. arXiv:1611.02903 (2016)
Wright, J., Ganesh, A., Min, K., Ma, Y.: Compressive principal component pursuit. Information and Inference: A Journal of the IMA 2(1), 32–68 (2013)
Chambolle, A., Pock, T.: An introduction to continuous optimization for imaging. Acta Numerica 25, 161–319 (2016)
Sra, S., Nowozin, S., Wright, S.J.: Optimization for machine learning. Mit Press (2012)
Gonċalves, M. L. N., Melo, J.G., Monteiro, R.D.C.: Convergence rate bounds for a proximal ADMM with over-relaxation stepsize parameter for solving nonconvex linearly constrained problems. arXiv:1702.01850 (2017)
Gu, Y, Jiang, B, Han, D.: A semi-proximal-based strictly contractive Peaceman-Rachford splitting method. arXiv:1506.02221 (2015)
Gonçalves, M.L.N.: On the pointwise iteration-complexity of a dynamic regularized ADMM with over-relaxation stepsize. Appl. Math. Comput. 336, 315–325 (2018)
Gonçalves, M.L.N., Alves, M.M., Melo, J.G.: Pointwise and ergodic convergence rates of a variable metric proximal alternating direction method of multipliers. J. Optim. Theory Appl. 177(2), 448–478 (2018)
Facchinei, F, Pang, J S: Finite-Dimensional Variational Inequalities and Complementarity Problems, vol. 1, Springer Series in Operations Research. Springer, New York (2003)
He, B., Yuan, X.: On the O(1/n) convergence rate of the Douglas-Rachford alternating direction method. SIAM J. Num. Anal. 50(2), 700–709 (2012)
Tibshirani, R.: Regression shrinkage and selection via the lasso. J. Royal Stat. Soc., Series B. 58(1), 267–288 (1996)
He, B., Ma, F., Yuan, X.: Optimal proximal augmented Lagrangian method and its application to full Jacobian splitting for multi-block separable convex minimization problems. IMA J. Num. Anal. https://doi.org/10.1093/imanum/dry092 (2019)
He, B., Ma, F., Yuan, X.: Optimally linearizing the alternating direction method of multipliers for convex programming. Comput. Optim. Appl. https://doi.org/10.1007/s10589-019-00152-3 (2017)
Gao, B., Ma, F.: Symmetric alternating direction method with indefinite proximal regularization for linearly constrained convex optimization. J. Optim. Theory Appl. 176(1), 178–204 (2018)
Acknowledgments
The author is grateful to the anonymous referees and the editor for their valuable comments and suggestions which have helped us improve the presentation of this paper. He would like to thank Professor Bingsheng He for fruitful discussions and suggestions regarding this project and thank Professor Shiqian Ma for providing the SPCP codes.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This author was supported by the NSFC Grant 11701564 and 11871029.
Rights and permissions
About this article
Cite this article
Ma, F. Convergence study on the proximal alternating direction method with larger step size. Numer Algor 85, 399–425 (2020). https://doi.org/10.1007/s11075-019-00819-2
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11075-019-00819-2
Keywords
- Alternating direction method of multipliers
- Convex programming
- Proximal regularization
- Convergence analysis