Abstract
The augmented Lagrangian method (ALM) is one of the most useful methods for constrained optimization. Its convergence has been well established under convexity assumptions or smoothness assumptions, or under both assumptions. ALM may experience oscillations and divergence when the underlying problem is simultaneously nonconvex and nonsmooth. In this paper, we consider the linearly constrained problem with a nonconvex (in particular, weakly convex) and nonsmooth objective. We modify ALM to use a Moreau envelope of the augmented Lagrangian and establish its convergence under conditions that are weaker than those in the literature. We call it the Moreau envelope augmented Lagrangian (MEAL) method. We also show that the iteration complexity of MEAL is \(o(\varepsilon ^{-2})\) to yield an \(\varepsilon \)-accurate first-order stationary point. We establish its whole sequence convergence (regardless of the initial guess) and a rate when a Kurdyka–Łojasiewicz property is assumed. Moreover, when the subproblem of MEAL has no closed-form solution and is difficult to solve, we propose two practical variants of MEAL, an inexact version called iMEAL with an approximate proximal update, and a linearized version called LiMEAL for the constrained problem with a composite objective. Their convergence is also established.
Similar content being viewed by others
Availability of data and materials
Enquiries about data availability should be directed to the authors.
Notes
Locally linear convergence means exponentially fast convergence to a local minimum from a sufficiently close initial point.
References
Andreani, R., Birgin, E.G., Martinez, J.M., Schuverdt, M.L.: On augmented Lagrangian methods with general lower-level constraints. SIAM J. Optim. 18(4), 1286–1309 (2007)
Andreani, R., Birgin, E.G., Martinez, J.M., Schuverdt, M.L.: Augmented Lagrangian methods under the constant positive linear dependence constraint qualification. Math. Program. 111, 5–32 (2008)
Andreani, R., Birgin, E.G., Martinez, J.M., Schuverdt, M.L.: Second-order negative-curvature methods for box-constrained and general constrained optimization. Comput. Optim. Appl. 45(2), 209–236 (2010)
Andreani, R., Fazzio, N., Schuverdt, M.L., Secchin, L.: A sequential optimality condition related to the quasi-normality constraint qualification and its algorithmic consequences. SIAM J. Optim. 29(1), 743–766 (2019)
Andreani, R., Secchin, L., Silva, P.: Convergence properties of a second order augmented Lagrangian method for mathematical programs with complementarity constraints. SIAM J. Optim. 28(3), 2574–2600 (2018)
Armand, P., Omheni, R.: A globally and quadratically convergent primal-dual augmented Lagrangian algorithm for equality constrained optimization. Optim. Methods Softw. 32(1), 1–21 (2017)
Attouch, H., Bolte, J.: On the convergence of the proximal algorithm for nonsmooth functions involving analytic features. Math. Program. 116, 5–16 (2009)
Attouch, H., Bolte, J., Svaiter, B.F.: Convergence of descent methods for semi-algebraic and tame problems: proximal algorithms, forward-backward splitting, and regularized gauss-seidel methods. Math. Program. 137, 91–219 (2013)
Bertsekas, D.P.: In: Convergence rate of penalty and multiplier methods, pp. 260–264. SanDiego, California (1973)
Bertsekas, D.P.: On penalty and multiplier methods for constrained minimization. SIAM J. Control. Optim. 14(2), 216–235 (1976)
Bertsekas, D.P.: Constrained Optimization and Lagrange Multiplier Methods. Academic Press, London (1982)
Bian, W., Chen, X., Ye, Y.: Complexity analysis of interior point algorithms for non-Lipschitz and nonconvex minimization. Math. Program. 149(1), 301–327 (2005)
Birgin, E.G., Castillo, R., Martinez, J.M.: Numerical comparison of augmented Lagrangian algorithms for nonconvex problems. Comput. Optim. Appl. 31, 31–56 (2005)
Birgin, E.G., Floudas, C.A., Martinez, J.M.: Global minimization using an augmented Lagrangian method with variable lower-level constraints. Math. Program. 125, 139–162 (2010)
Birgin, E.G., Floudas, C.A., Martinez, J.M.: The boundedness of penalty parameters in an augmented Lagrangian method with constrained subproblems. Optim. Methods Softw. 27(6), 1001–1024 (2012)
Birgin, E.G., Haeser, G., Ramos, A.: Augmented Lagrangians with constrained subproblems and convergence to second-order stationary points. Comput. Optim. Appl. 69(1), 51–75 (2018)
Birgin, E.G., Martinez, J.M.: Practical Augmented Lagrangian Methods for Constrained Optimization, vol. 10. SIAM, Philadelphia (2014)
Birgin, E.G., Martinez, J.M.: Complexity and performance of an augmented Lagrangian algorithm. Optim, Methods Softw (2020)
Bochnak, J., Coste, M., Roy, M.F.: Real Algebraic Geometry, vol. 36. Springer, Berlin (1998)
Bolte, J., Daniilidis, A., Lewis, A.: The łojasiewicz inequality for nonsmooth subanalytic functions with applications to subgradient dynamical systems. SIAM J. Optim. 17(4), 1205–1223 (2007)
Bolte, J., Daniilidis, A., Lewis, A., Shiota, M.: Clark subgradients of stratifiable functions. SIAM J. Optim. 18(2), 556–572 (2007)
Bolte, J., Sabach, S., Teboulle, M.: Proximal alternating linearized minimization for nonconvex and nonsmooth problems. Math. Program. 146(1), 459–494 (2014)
Boyd, S., Vandenberghe, L.: Convex Optimization. Cambridge University Press, Cambridge (2004)
Conn, A.R., Gould, N.I.M., Startenaer, A., Toint, P.L.: Convergence properties of an augmented Lagrangian algorithm for optimization with a combination of general equality and linear constraints. SIAM J. Optim. 6, 674–703 (1996)
Conn, A.R., Gould, N.I.M., Toint, P.L.: A globally convergent augmented Lagrangian algorithm for optimization with general constraints and simple bounds. SIAM J. Numer. Anal. 28, 545–572 (1991)
Conn, A.R., Gould, N.I.M., Toint, P.L.: Trust-Region Methods. SIAM, Philadelphia (2000)
Curtis, F.E., Jiang, H., Robinson, D.P.: An adaptive augmented Lagrangian method for large-scale constrained optimization. Math. Program. 152(1), 201–245 (2015)
Davis, D., Drusvyatskiy, D.: Stochastic model-based minimization of weakly convex functions. SIAM J. Optim. 29(1), 207–239 (2019)
Deng, W., Lai, M.J., Peng, Z., Yin, W.: Parallel multi-block admm with \(o(1/k)\) convergence. J. Sci. Comput. 71, 712–736 (2017)
Drusvyatskiy, D.: The proximal point method revisited. SIAG/OPT Views and News 26, 1–8 (2018)
Drusvyatskiy, D., Paquette, C.: Efficiency of minimizing compositions of convex functions and smooth maps. Math. Program. 178, 503–558 (2019)
Fan, J., Li, R.: Variable selection via nonconcave penalized likelihood and its oracle properties. J. Am. Stat. Assoc. 96, 1348–1360 (2001)
Fernadez, D., Solodov, M.V.: Local convergence of exact and inexact augmented Lagrangian methods under the second-order sufficient optimality condition. SIAM J. Optim. 22(2), 384–407 (2012)
Grapiglia, G.N., Yuan, Y.X.: On the complexity of an augmented Lagrangian method for nonconvex optimization. ArXiv e-prints (2019)
Haeser, G., Liu, H., Ye, Y.: Optimality condition and complexity analysis for linearly-constrained optimization without differentiability on the boundary. Math. Program. 178, 263–299 (2019)
Hajinezhad, D., Hong, M.: Perturbed proximal primal-dual algorithm for nonconvex nonsmooth optimization. Math. Program. 176, 207–245 (2019)
Hestenes, M.R.: Multiplier and gradient methods. J. Optim. Theory Appl. 4, 303–320 (1969)
Hong, M., Hajinezhad, D., Zhao, M.M.: Prox-pda,: In: The proximal primal-dual algorithm for fast dostributed nonconvex optimization and learning over networks, pp. 1529–1538. , Sydney, Australia (2017)
Jiang, B., Lin, T., Ma, S., Zhang, S.: Structured nonconvex and nonsmooth otpmization: algorithms and iteration complexity analysis. Comput. Optim. Appl. 72(1), 115–157 (2019)
Krantz, S., Parks, H.R.: A Primer of Real Analytic Functions, 2nd edn. Birkhauser, Basel (2002)
Kurdyka, K.: On gradients of functions definable in o-minimal structures. Annales de l’institut Fourier 48(3), 769–783 (1998)
Li, G., Pong, T.K.: Calculus of the exponent of Kurdyka–łojasiewicz inequality and its applications to linear convergence of first-order methods. Found. Comput. Math. 18, 1199–1232 (2018)
Łojasiewicz, S.: Une propriété topologique des sous-ensembles analytiques réels. In: Les Équations aux dérivées partielles. Éditions du centre National de la Recherche Scientifique, Paris pp. 87–89 (1963)
Łojasiewicz, S.: Sur la geometrie semi-et sous-analytique. Annales de l’institut Fourier 43(5), 1575–1595 (1993)
Mordukhovich, B.S.: Variational Analysis and Generalized Differentiation I: Basic Theory. Springer-Verlag, New York (2006)
Moreau, J.: Proximité et dualité dans un espace hilbertien. Bull. Soc. Math. France 93, 273–299 (1965)
Nocedal, J., Wright, S.J.: Numerical Optimization. Springer-Verlag, New York (1999)
Nouiehed, M., Lee, J.D., Razaviyayn, M.: Convergence to second-order stationary for constrained non-convex optimization. ArXiv e-prints (2018)
Nurminskii, E.A.: The quasigradient method for the solving of the nonlinear programming problems. Cybernetics 9, 145–150 (1973)
O’Neill, M., Wright, S.J.: A log-barrier newton-cg method for bound constrained optimization with complexity guarantees. IMA J. Numer. Anal. 1–38 (2020)
Polyak, B.T., Tretyakov, N.V.: The method of penalty bounds for constrained extremum problems. Zh. Vych Mat i Mat. Fiz, 13:34–46 = U.S.S.R. Computational Mathematics and Mathmatical. Physics 13, 42–58 (1973)
Powell, M.J.D.: In: Optimization, R. Fletcher. (ed.) A method for nonlinear constraints in minimization problems, pp. 283–298. Academic Press, London (1969)
Rockafellar, R.T.: The multiplier method of hestenes and powell applied to convex programming. J. Optim. Theory Appl. 12, 555–562 (1973)
Rockafellar, R.T.: Augmented Lagrangians and applications of the proximal point algorithm in convex programming. Math. Oper. Res. 1(2), 97–116 (1976)
Rockafellar, R.T., Wets, R.J.B.: Variational Analysis. Springer, New York (1997)
Shiota, M.: Geometry of Subanalytic and Semialgebraic Sets (Progress in Mathematics). Birkhauser, Basel (1997)
Tretykov, N.Y.: The method of penalty estimates of convex programming. Econ. Math. Methods (Russian) 9, 525–540 (1973)
Wang, Y., Yin, W., Zeng, J.: Global convergence of admm in nonconvex nonsmooth optimization. J. Sci. Comput. 78, 29–63 (2019)
Xie, Y., Wright, S.J.: Complexity of proximal augmented Lagrangian for nonconvex optimalization with nonlinear equality constraints. ArXiv e-prints (2019)
Xu, Y., Yin, W.: A block coordinate descent method for regularized multiconvex optimizaton with applications to nonnegative tensor factorization and completion. SIAM J. Imaging Sci. 6(3), 1758–1789 (2013)
Yu, P., Li, G., Pong, T.: Kurdyka–łojasiewicz exponent via inf-projection. Found. Comput. Math. (2021). https://doi.org/10.1007/s10208-021-09528-6
Zeng, J., Lau, T.T.K., Lin, S.B., Yao, Y.: In: Global convergence of block coordinate descent in deep learning. , Long Beach, California (2019) . (PMLR 97)
Zeng, J., Lin, S.B., Yao, Y., Zhou, D.X.: On admm in deep learning: convergence and saturation-avoidance. J. Mach. Learn. Res. 22(199), 1–67 (2021)
Zeng, J., Yin, W.: On nonconvex descentralized gradient descent. IEEE Trans. Signal Process. 66(11), 2834–2848 (2018)
Zhang, C.H.: Nearly unbiased variable selection under minimax concave penalty. Ann. Stat. 38(2), 894–942 (2010)
Zhang, J., Luo, Z.Q.: A global dual error bound and its application to the analysis of linearly constrained nonconvex optimization. ArXiv e-prints (2020)
Zhang, J., Luo, Z.Q.: A proximal alternating direction method of multiplier for linearly constrained nonconvex minimization. SIAM J. Optim. 30(3), 2272–2302 (2020)
Zhou, D.X.: Universality of deep convolutional neural networks. Appl. Comput. Harmonic Anal. 48, 787–794 (2020)
Acknowledgements
The authors have not disclosed any funding.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
We thank Kaizhao Sun for discussions that help us complete this paper, as well as presenting to us an additional approach to ensure boundedness. The work of J. Zeng is partly supported by National Natural Science Foundation of China (No. 61977038) and the Thousand Talents Plan of Jiangxi Province (No. jxsq2019201124). The work of D.-X. Zhou is partly supported by Research Grants Council of Hong Kong (No. CityU 11307319), Laboratory for AI-powered Financial Technologies, and the Hong Kong Institute for Data Science.
Rights and permissions
About this article
Cite this article
Zeng, J., Yin, W. & Zhou, DX. Moreau Envelope Augmented Lagrangian Method for Nonconvex Optimization with Linear Constraints. J Sci Comput 91, 61 (2022). https://doi.org/10.1007/s10915-022-01815-w
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s10915-022-01815-w