Abstract
The problem of minimizing the rank of a symmetric positive semidefinite matrix subject to constraints can be cast equivalently as a semidefinite program with complementarity constraints (SDCMPCC). The formulation requires two positive semidefinite matrices to be complementary. This is a continuous and nonconvex reformulation of the rank minimization problem. We investigate calmness of locally optimal solutions to the SDCMPCC formulation and hence show that any locally optimal solution is a KKT point. We develop a penalty formulation of the problem. We present calmness results for locally optimal solutions to the penalty formulation. We also develop a proximal alternating linearized minimization (PALM) scheme for the penalty formulation, and investigate the incorporation of a momentum term into the algorithm. Computational results are presented.
Similar content being viewed by others
References
Alfakih, A.Y., Anjos, M.F., Piccialli, V., Wolkowicz, H.: Euclidean distance matrices, semidefinite programming, and sensor network localization. Port. Math. 68, 53–102 (2011)
Bai, L., Mitchell, J.E., Pang, J.: On conic QPCCs, conic QCQPs and completely positive programs. Math. Program. 159, 109–136 (2016)
Bolte, J., Sabach, S., Teboulle, M.: Proximal alternating linearized minimization for nonconvex and nonsmooth problems. Math. Program. 146, 459–494 (2014)
Burdakov, O., Kanzow, C., Schwartz, A.: Mathematical programs with cardinality constraints: reformulation by complementarity-type conditions and a regularization method. SIAM J. Optim. 26, 397–425 (2016)
Cai, J.-F., Candès, E.J., Shen, Z.: A singular value thresholding algorithm for matrix completion. SIAM J. Optim. 20, 1956–1982 (2010)
Cambier, L., Absil, P.-A.: Robust low-rank matrix completion by Riemannian optimization. SIAM J. Sci. Comput. 38, S440–S460 (2016)
Clarke, F.H.: Optimization and Nonsmooth Analysis. SIAM, Philadelphia (1990)
Ding, C., Sun, D., Ye, J.: First order optimality conditions for mathematical programs with semidefinite cone complementarity constraints. Math. Program. 147, 539–579 (2014)
Ding, Y., Krislock, N., Qian, J., Wolkowicz, H.: Sensor network localization, euclidean distance matrix completions, and graph realization. Optim. Eng. 11, 45–66 (2008)
Fazel, M., Hindi, H., Boyd, S.P.: A rank minimization heuristic with application to minimum order system approximation. In: Proceedings of the 2001 American Control Conference, June 2001, IEEE, pp. 4734–4739 (2001). https://faculty.washington.edu/mfazel/nucnorm.html
Feng, M., Mitchell, J.E., Pang, J., Shen, X., Wächter, A.: Complementarity formulations of \(\ell \)0-norm opti-mization problems. Tech. Report, Department of Mathematical Sciences, Rensselaer Polytechnic Institute, Troy, NY (2013). Revised: May 2016
Ghadimi, S., Lan, G.: Accelerated gradient methods for nonconvex nonlinear and stochastic programming. Math. Program. 156, 59–99 (2016)
Grant, M., Boyd, S., Ye, Y.: Disciplined convex programming, in global optimization: from theory to implementation. In: Liberti, L., Maculan, N. (eds.) Nonconvex Optimization and its Applications, pp. 155–210. Springer, Berlin (2006)
Hestenes, M.R., Stiefel, E.: Methods of Conjugate Gradients for Solving Linear Systems. NBS, Bikaner (1952)
Hsieh, C.-J., Olsen, P.: Nuclear norm minimization via active subspace selection. In: International Conference Machine Learning Proceedings, Beijing, China, pp. 575–583 (2014)
Huang, X.X., Teo, K.L., Yang, X.Q.: Calmness and exact penalization in vector optimization with cone constraints. Comput. Optim. Appl. 35, 47–67 (2006)
Keshavan, R.H., Montanari, A., Oh, S.: Matrix completion from a few entries. IEEE Trans. Inf. Theory 56, 2980–2998 (2010)
Kocvara, M., Stingl, M.: PENNON—a code for convex nonlinear and semidefinite programming. Optim. Methods Softw. 18, 317–333 (2003)
Krislock, N., Wolkowicz, H.: Euclidean distance matrices and applications. In: Anjos, M., Lasserre, J. (eds.) Handbook on Semidefinite, Conic and Polynomial Optimization, International Series in Operational Research and Management Science. Springer, pp. 879–914 (2012)
Laurent, M.: A tour d’horizon on positive semidefinite and Euclidean distance matrix completion problems. In: Topics in Semidefinite and Interior Point Methods, vol. 8 of The Fields Institute for Research in Mathematical Sciences, Communications Series, AMS, Providence, RI (1998)
Li, H., Lin, Z.: Accelerated proximal gradient methods for nonconvex programming. In: Advances in Neural Information Processing Systems, Montreal, Canada, pp. 379–387 (2015)
Li, Q., Qi, H.-D.: A sequential semismooth Newton method for the nearest low-rank correlation matrix problem. SIAM J. Optim. 21, 1641–1666 (2011)
Li, X., Ling, S., Strohmer, T., Wei, K.: Rapid, robust, and reliable blind deconvolution via nonconvex optimization. Tech. Report, Department of Mathematics, University of California Davis, Davis, CA 95616 (2016)
Lin, Q., Lu, Z., Xiao, L.: An accelerated randomized proximal coordinate gradient method and its application to regularized empirical risk minimization. SIAM J. Optim. 25, 2244–2273 (2015)
Lin, Z., Chen, M., Ma, Y.: The augmented Lagrange multiplier method for exact recovery of corrupted low-rank matrices. Tech. Report, Perception and Decision Lab, University of Illinois, Urbana-Champaign, IL (2010)
Liu, Y.-J., Sun, D., Toh, K.-C.: An implementable proximal point algorithmic framework for nuclear norm minimization. Math. Program. 133, 399–436 (2012)
Liu, Z., Vandenberghe, L.: Interior-point method for nuclear norm approximation with application to system identification. SIAM J. Matrix Anal. Appl. 31, 1235–1256 (2009)
Lu, S.: Relation between the constant rank and the relaxed constant rank constraint qualifications. Optimization 61, 555–566 (2012)
Luenberger, D.G.: Linear and Nonlinear Programming, 2nd edn. Addison-Wesley, Menlo Park (1984)
Ma, S., Goldfarb, D., Chen, L.: Fixed point and Bregman iterative methods for matrix rank minimization. Math. Program. 128, 321–353 (2011)
Mohan, K., Fazel, M.: Reweighted nuclear norm minimization with application to system identification. In: Proceedings of the American Control Conference, Baltimore, MD. IEEE, pp. 2953–2959 (2010)
Mohan, K., Fazel, M.: Iterative reweighted algorithms for matrix rank minimization. Journal of Machine Learning Research 13, 3441–3473 (2012)
Nesterov, Y.: Introductory Lectures on Convex Optimization: A Basic Course. Springer, New York (2013)
Nesterov, Y.E.: A method for unconstrained convex minimization problem with the rate of convergence \(o(1/k^2)\). In: Doklady AN SSSR, vol 269, pp. 543–547 (1983). translated as Soviet Mathematics Doklady
Nesterov, Y.E.: How to advance in structural convex optimization. Optima 78, 2–5 (2008)
Pong, T.K., Tseng, P.: (Robust) edge-based semidefinite programming relaxation of sensor network localization. Math. Program. 130, 321–358 (2011)
Qi, H., Sun, D.: A quadratically convergent Newton method for computing the nearest correlation matrix. SIAM J. Matrix Anal. Appl. 28, 360–385 (2006)
Recht, B., Fazel, M., Parrilo, P.A.: Guaranteed minimum-rank solutions of linear matrix inequalities via nuclear norm minimization. SIAM Rev. 52, 471–501 (2010)
Schmidt, M., Roux, N.L., Bach, F.R.: Convergence rates of inexact proximal-gradient methods for convex optimization. In: Advances in Neural Information Processing Systems, Granada, Spain, pp. 1458–1466 (2011)
So, A.M.-C., Ye, Y.: Theory of semidefinite programming for sensor network localization. Math. Program. 109, 367–384 (2007)
Srebro, N., Jaakkola, T.: Weighted low-rank approximations. In: Proceedings of the International Conference on Machine Learning, Atlanta, GA, pp. 720–727 (2003)
Sun, R., Luo, Z.-Q., Guaranteed matrix completion via nonconvex factorization. In: 2015 IEEE 56th Annual Symposium on Foundations of Computer Science, pp. 270–289 (2015)
Tanner, J., Wei, K.: Low rank matrix completion by alternating steepest descent methods. Appl. Comput. Harmonic Anal. 40, 417–429 (2016)
Toh, K.C., Yun, S.: An accelerated proximal gradient algorithm for nuclear norm regularized linear least squares problems. Pac. J. Optim. 6, 615–640 (2010)
Vandereycken, B.: Low-rank matrix completion by Riemannian optimization. SIAM J. Optim. 23, 1214–1236 (2013)
Wei, K., Cai, J.-F., Chan, T.F., Leung, S.: Guarantees of Reimannian optimization for low rank matrix completion. Tech. Report, Department of Mathematics, University of California at Davis, CA, USA (2016)
Wu, J., Zhang, L.: On properties of the bilinear penalty function method for mathematical programs with semidefinite cone complementarity constraints. Set Valued Var. Anal. 23, 277–294 (2015)
Yamashita, H., Yabe, H.: A survey of numerical methods for nonlinear semidefinite programming. J. Oper. Res. Soc. Jpn. 58, 24–60 (2015)
Ye, J.J., Zhou, J.: First-order optimality conditions for mathematical programs with second-order cone complementarity constraints. Tech. Report, Department of Mathematics and Statistics, University of Victoria, Victoria, BC, Canada (2015)
Zhai, J., Huang, X.X.: Calmness and exact penalization in vector optimization under nonlinear perturbations. J. Optim. Theory Appl. 162, 856–872 (2014)
Zhang, H., Lin, Z., Zhang, C.: A Counterexample for the Validity of Using Nuclear Norm as a Convex Surrogate of Rank, in Machine Learning and Knowledge Discovery in Databases, pp. 226–241. Springer, Berlin (2013)
Zhang, Y., Zhang, L., Wu, J.: Convergence properties of a smoothing approach for mathematical programs with second-order cone complementarity constraints. Set Valued Var. Anal. 19, 609–646 (2011)
Author information
Authors and Affiliations
Corresponding author
Additional information
This work was supported in part by the Air Force Office of Sponsored Research under Grants FA9550-08-1-0081 and FA9550-11-1-0260 and by the National Science Foundation under Grants Numbers CMMI-1334327 and DMS-1736326.
Rights and permissions
About this article
Cite this article
Shen, X., Mitchell, J.E. A penalty method for rank minimization problems in symmetric matrices. Comput Optim Appl 71, 353–380 (2018). https://doi.org/10.1007/s10589-018-0010-6
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10589-018-0010-6