Abstract
In this paper, we consider the augmented Lagrangian duality for optimization problems with sparsity and abstract set constraints and present second-order conditions for the existence of augmented Lagrange multipliers by calculating the second-order epi-derivative of the augmented Lagrangian. The ingredient of the augmented Lagrangian here includes the indicator function of a sparse set and a composition of the Moreau envelope of the indicator function of a second-order regular set and a twice continuously differentiable mapping. The main process depends heavily on the calculation of the second-order epi-derivative of the indicator function of sparse set which is shown to be second-order regular and also parabolically regular. The second-order sufficient conditions for the sparse nonlinear programming, the sparse inverse covariance selection problem, and the sparse second-order cone programming are obtained as special cases of our general results. We prove that the existence of augmented Lagrange multipliers ensures the exactness of penalty functions and the stability of augmented solutions under small perturbations of the corresponding augmented Lagrange multipliers.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Beck, A., Eldar, Y.: Sparsity constrained nonlinear optimization: optimality conditions and algorithms. SIAM J. Optim. 23, 1480–1509 (2013)
Beck, A., Hallak, N.: On the minimization over sparse symmetric sets: projections, optimality conditions, and algorithm. Math. Oper. Res. 41, 196–223 (2015)
Bonnans, J.F., Shapiro, A.: Perturbation Analysis of Optimization Problems. Springer, New York (2000)
Combettes, P.L., Wajs, V.R.: Signal recovery by proximal forward-backward splitting. Multiscale Model. Simul. 4, 1168–1200 (2005)
Eberhard, A., Wenczel, R.: Some sufficient optimality conditions in nonsmooth analysis. SIAM J. Optim. 20, 251–296 (2009)
Fernández, D., Solodov, M.V.: Local convergence of exact and inexact augmented Lagrangian methods under the second-order sufficient optimality condition. SIAM J. Optim. 22, 384–407 (2012)
Kan, C., Song, W.: Augmented Lagrangian duality for composite optimization problems. J. Optim. Theory Appl. 165, 763–784 (2015)
Kan, C., Song, W.: Second-order conditions for existence of augmented Lagrange multipliers for eigenvalue composite optimization problems. J. Glob. Optim. 63, 77–97 (2015)
Lu, Z.: Optimization over sparse symmetric sets via a nonmonotone projected gradient method. https://doi.org/10.48550/arXiv.1509.08581
Lu, Z., Zhang, Y.: Sparse approximation via penalty decomposition methods. SIAM J. Optim. 23, 2448–2478 (2013)
Mohammadi, A., Mordukhovich, B., Sarabi, M.: Parabolic regularity in geometric variational analysis. Trans. Am. Math. Soc. 374, 1711–1763 (2021)
Pan, L., Xiu, N., Zhou, S.: On solusions of sparsity constrained optimization. J. Oper. Res. Soc. China. 4, 421–439 (2015)
Pan, L., Xiu, N., Fan, J.: Optimality conditions for sparse nonlinear programming. Sci. China Math. 60, 759–776 (2017)
Pan, L., Luo, Z., Xiu, N.: Restricted Robinson consraint qualification and optimality for cardinality-constrained cone programming. J. Optim. Theory Appl. 175, 104–118 (2017)
Rockafellar, R.T.: Augmented Lagrange multiplier functions and duality in nonconvex programming. SIAM J. Control. 12, 268–285 (1974)
Rockafellar, R.T.: First- and second-order epi-differentiability in nonlinear programming. Trans. Am. Math. Soc. 307, 75–108 (1988)
Rockafellar, R.T.: Second-order optimality conditions in nonlinear programming obtained by way of epi-derivatives. Math. Oper. Res. 14, 462–484 (1989)
Rockafellar, R.T., Wets, R.J.-B.: Variational Analysis. Springer, New York (1998)
Shapiro, A., Sun, J.: Some properties of the augmented Lagrangian in cone constrained optimization. Math. Oper. Res. 29, 479–491 (2004)
Acknowledgements
The research of the first author was partly supported by the National Natural Sciences Grant of China (No. 11701126) and the Foundation of Heilongjiang Department of Education (2017-KYYWF-0136 and UNPYSCT-2018181). The research of the second author was supported by the National Natural Sciences Grant of China (No. 11871182).
Author information
Authors and Affiliations
Corresponding author
Additional information
Communicated by Ebrahim Sarabi.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Kan, C., Song, W. Second-Order Conditions for the Existence of Augmented Lagrange Multipliers for Sparse Optimization. J Optim Theory Appl 201, 103–129 (2024). https://doi.org/10.1007/s10957-024-02382-w
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10957-024-02382-w
Keywords
- Augmented Lagrange multiplier
- Augmented Lagrangian duality
- Exact penalty
- Second-order epi-derivative
- Sparsity