Abstract
In this paper, we consider a least square semidefinite programming problem under ellipsoidal data uncertainty. We show that the robustification of this uncertain problem can be reformulated as a semidefinite linear programming problem with an additional second-order cone constraint. We then provide an explicit quantitative sensitivity analysis on how the solution under the robustification depends on the size/shape of the ellipsoidal data uncertainty set. Next, we prove that, under suitable constraint qualifications, the reformulation has zero duality gap with its dual problem, even when the primal problem itself is infeasible. The dual problem is equivalent to minimizing a smooth objective function over the Cartesian product of second-order cones and the Euclidean space, which is easy to project onto. Thus, we propose a simple variant of the spectral projected gradient method (Birgin et al. in SIAM J. Optim. 10:1196–1211, 2000) to solve the dual problem. While it is well-known that any accumulation point of the sequence generated from the algorithm is a dual optimal solution, we show in addition that the dual objective value along the sequence generated converges to a finite value if and only if the primal problem is feasible, again under suitable constraint qualifications. This latter fact leads to a simple certificate for primal infeasibility in situations when the primal feasible set lies in a known compact set. As an application, we consider robust correlation stress testing where data uncertainty arises due to untimely recording of portfolio holdings. In our computational experiments on this particular application, our algorithm performs reasonably well on medium-sized problems for real data when finding the optimal solution (if exists) or identifying primal infeasibility, and usually outperforms the standard interior-point solver SDPT3 in terms of CPU time.
Similar content being viewed by others
Notes
We remark that many popular methods such as Newton type method described in [17], accelerated proximal gradient methods (see, for example, [25–28, 37]) and the alternating direction method of multipliers (see, for example, [10, 12, 14, 15, 18, 19, 41]), could also be suitably adapted to solve (DSDP) when (RSDP) is feasible. However, it is not immediate to us how primal infeasibility can be readily certified in these algorithms.
We have chosen SDPT3 since it is an off-the-shelf SDP solver and implements a second-order method that returns solutions with high accuracy; this latter point is important for benchmark purpose and for understanding the behavior of a new model.
References
Ben-Tal, A., Nemirovski, A.: Robust solutions of linear programming problems contaminated with uncertain data. Math. Program. 88, 411–424 (2000)
Ben-Tal, A., Nemirovski, A.: Selected topics in robust convex optimization. Math. Program. 112, 125–158 (2008)
Ben-Tal, A., El Ghaoui, L., Nemirovski, A.: Robust Optimization. Princeton Series in Applied Mathematics. Princeton University Press, Princeton (2009)
Berkowitz, J.: A coherent framework for stress testing. J. Risk 2(2), 5–15 (1999)
Bertsimas, D., Brown, D.B.: Constructing uncertainty sets for robust linear optimization. Oper. Res. 57, 1483–1495 (2009)
Bertsimas, D., Brown, D.B., Caramanis, C.: Theory and applications of robust optimization. SIAM Rev. 53, 464–501 (2011)
Birge, J.R., Louveaux, F.: Introduction to Stochastic Programming. Springer, Berlin (1997)
Birgin, E.G., Martínez, J.M., Raydan, M.: Nonmonotone spectral projected gradient methods on convex sets. SIAM J. Optim. 10, 1196–1211 (2000)
Chou, C.-C., Ng, K.-F., Pang, J.-S.: Minimizing and stationary sequences of constrained optimization problems. SIAM J. Control Optim. 36, 1908–1936 (1998)
Eckstein, J., Bertsekas, D.P.: On the Douglas-Rachford splitting method and the proximal point algorithm for maximal monotone operators. Math. Prog. 55, 293–318 (1992)
El Ghaoui, L., Lebret, H.: Robust solutions to least-squares problems with uncertain data. SIAM J. Matrix Anal. Appl. 18, 1035–1064 (1997)
Fortin, M., Glowinski, R.: On decomposition-coordination methods using an augmented Lagrangian. In: Fortin, M., Glowinski, R. (eds.) Augmented Lagrangion Methods: Applications to the Solution of Boundary Problems. North-Holland, Amsterdam (1983)
Fukushima, M., Luo, Z.-Q., Tseng, P.: Smoothing functions for second-order-cone complementarity problems. SIAM J. Optim. 12, 436–460 (2001)
Gabay, D.: Applications of the method of multipliers to variational inequalities. In: Fortin, M., Glowinski, R. (eds.) Augmented Lagrangion Methods: Applications to the Solution of Boundary Problems. North-Holland, Amsterdam (1983)
Gabay, D., Mercier, B.: A dual algorithm for the solution of nonlinear variational problems via finite element approximations. Comput. Math. Appl. 2, 17–40 (1976)
Gafni, E.M., Bertsekas, D.P.: Two metric projection methods for constrained optimization. SIAM J. Control Optim. 22, 936–964 (1984)
Gao, Y., Sun, D.: Calibrating least squares semidefinite programming with equality and inequality constraints. SIAM J. Matrix Anal. Appl. 31, 1432–1457 (2009)
Glowinski, R., Marroco, A.: Sur l’approximation, par elements finis d’ordre un, et la resolution, par penalisation-dualit’e, d’une classe de problemes de Dirichlet non lineares. Revue Francaise d’Automatique Informatique et Recherche Op’erationelle 9(r–2), 41–76 (1975)
He, B., Liao, L., Han, D., Yang, H.: A new inexact alternating directions method for monotone variational inequalities. Math. Program. 92, 103–118 (2002)
Hu, H.: Geometric condition measures and smoothness condition measures for closed convex sets and linear regularity of infinitely many closed convex sets. J. Optim. Theory Appl. 126, 287–308 (2005)
Kim, J., Finger, C.C.: A stress test to incorporate correlation breakdown. J. Risk 2(3), 5–19 (2000)
Kupiec, P.H.: Stress testing in a value at risk framework. J. Deriv. 6, 7–24 (1998)
Li, C., Ng, K.F., Pong, T.K.: The SECQ, linear regularity, and the strong CHIP for an infinite system of closed convex sets in normed linear spaces. SIAM J. Optim. 18, 643–665 (2007)
Lu, Z., Zhang, Y.: An augmented Lagrangian approach for sparse principal component analysis. Math. Program. 135, 149–193 (2012)
Nesterov, Y.: A method for solving a convex programming problem with convergence rate O(1/k 2). Sov. Math. Dokl. 27(2), 372–376 (1983)
Nesterov, Y.: Introductory Lectures on Convex Optimization: A Basic Course. Kluwer Academic, Amsterdam (2003)
Nesterov, Y.: Excessive gap technique in nonsmooth convex minimization. SIAM J. Optim. 16, 235–249 (2005)
Nesterov, Y.: Smooth minimization of non-smooth functions. Math. Program. 103, 127–152 (2005)
Qi, H., Sun, D.: Correlation stress testing for value-at-risk: an unconstrained convex optimization approach. Comput. Optim. Appl. 45, 427–462 (2010)
Rebonato, R., Jäckel, P.: The most general methodology for creating a valid correlation matrix for risk management and option pricing purposes. J. Risk 2(2), 17–27 (1999)
Robinson, S.M.: An application of error bounds for convex programming in a linear space. SIAM J. Control 13, 271–273 (1975)
Rockafellar, R.T.: Convex Analysis. Princeton University Press, Princeton (1970)
Shapiro, A., Dentcheva, D., Ruszczynski, A.: Lectures on Stochastic Programming: Modeling and Theory. SIAM, Philadelphia (2009)
Soyster, A.L.: Convex programming with set-inclusive constraints and applications to inexact linear programming. Oper. Res. 21, 1154–1157 (1973)
Toh, K.C., Todd, M.J., Tütüncü, R.H.: SDPT3—a Matlab software package for semidefinite programming. Optim. Method Softw. 11, 545–581 (1999)
Toint, Ph.L.: Global convergence of a class of trust-region methods for nonconvex minimization in Hilbert space. IMA J. Numer. Anal. 8, 231–252 (1988)
Tseng, P.: Approximation accuracy, gradient methods, and error bound for structured convex optimization. Math. Program. 125, 263–295 (2010)
Tseng, P., Yun, S.: A coordinate gradient descent method for nonsmooth separable minimization. Math. Program. 117, 387–423 (2007)
Turkay, S., Epperlein, E., Christofides, N.: Correlation stress testing for value-at-risk. J. Risk 5(4), 75–89 (2003)
Tütüncü, R.H., Toh, K.C., Todd, M.J.: Solving semidefinite-quadratic-linear programs using SDPT3. Math. Program. Series B 95, 189–217 (2003)
Yang, J., Zhang, Y.: Alternating direction algorithms for ℓ 1-problems in compressive sensing. SIAM J. Sci. Comput. 33, 250–278 (2011)
Author information
Authors and Affiliations
Corresponding author
Additional information
G. Li author was partially supported by a research grant from Australian Research Council.
T.K. Pong author was supported by research grants from AFOSR and NSERC.
Appendix: Proof of Lemma 2.1
Appendix: Proof of Lemma 2.1
Proof
By a translation, we may assume without loss of generality that X 0=0. For a closed convex set Ω, let σ Ω denote the usual support function of Ω (that is, σ Ω (X)=sup Y∈Ω tr(YX) for all \(X \in {\mathcal{S}}^{n}\)) and let epiσ Ω be the epigraph of the support function. Now, notice that if \((Y_{1},\alpha_{1})\in {\mathrm{epi}}\,\sigma_{\varOmega_{1}}\) and \((Y_{2},\alpha_{2})\in {\mathrm{epi}}\,\sigma_{\varOmega_{2}}\) are such that \(\alpha_{1}+\alpha_{2}=\sigma_{\varOmega_{1}\cap \varOmega_{2}}(Y_{1}+Y_{2})\) and ∥Y 1+Y 2∥ F ≤1, then we have \(0\le \sigma_{\varOmega_{1}}(Y_{1})\le \alpha_{1}\) and \(\alpha_{2}\le \sigma_{\varOmega_{1}\cap \varOmega_{2}}(Y_{1}+Y_{2})\le R\). From these we obtain further that \(\delta\|Y_{2}\|_{F}\le \sigma_{\varOmega_{2}}(Y_{2})\le \alpha_{2}\le R\) and consequently
The proof of this lemma now follows similarly as in [23, Lemma 4.10], making use of the bounds derived above. □
Rights and permissions
About this article
Cite this article
Li, G., Ma, A.K.C. & Pong, T.K. Robust least square semidefinite programming with applications. Comput Optim Appl 58, 347–379 (2014). https://doi.org/10.1007/s10589-013-9634-8
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10589-013-9634-8