Abstract
Let B i be deterministic real symmetric m × m matrices, and ξ i be independent random scalars with zero mean and “of order of one” (e.g., \(\xi_{i}\sim \mathcal{N}(0,1)\)). We are interested to know under what conditions “typical norm” of the random matrix \(S_N = \sum_{i=1}^N\xi_{i}B_{i}\) is of order of 1. An evident necessary condition is \({\bf E}\{S_{N}^{2}\}\preceq O(1)I\), which, essentially, translates to \(\sum_{i=1}^{N}B_{i}^{2}\preceq I\); a natural conjecture is that the latter condition is sufficient as well. In the paper, we prove a relaxed version of this conjecture, specifically, that under the above condition the typical norm of S N is \(\leq O(1)m^{{1\over 6}}\): \({\rm Prob}\{||S_N||>\Omega m^{1/6}\}\leq O(1)\exp\{-O(1)\Omega^2\}\) for all Ω > 0 We outline some applications of this result, primarily in investigating the quality of semidefinite relaxations of a general quadratic optimization problem with orthogonality constraints \({\rm Opt} = \max\limits_{X_{j}\in{\bf R}^{m\times m}}\left\{F(X_1,\ldots ,X_k): X_jX_j^{\rm T}=I,\,j=1,\ldots ,k\right\}\), where F is quadratic in X = (X 1,... ,X k ). We show that when F is convex in every one of X j , a natural semidefinite relaxation of the problem is tight within a factor slowly growing with the size m of the matrices \(X_j : {\rm Opt}\leq {\rm Opt}(SDP)\leq O(1) [m^{1/3}+\ln k]{\rm Opt}\).
Similar content being viewed by others
References
Anstreicher K., Wolkowicz H. (2000) On Lagrangian relaxation of quadratic matrix constraints. SIAM J. Matrix Anal. Appl. 22, 41–55
Ben-Tal A., Nemirovski A., Roos C. (2002) Robust solutions of uncertain quadratic and conic-quadratic problems. SIAM J. Optim. 13, 535–560
Browne M.W. (1967) On oblique procrustes rotation. Psychometrika 32, 125–132
Edelman A., Arias T. Smith S.T. (1999) The geometry of algorithms with orthogonality constraints. SIAM J. Matrix Anal. Appl. 20, 303–353
Johnson, W.B., Schechtman, G.: Remarks on Talagrand’s deviation inequality for Rademacher functions. In: Odell, E., Rosenthal, H. (eds.) Functional Analysis (Austin, TX 1987/1989), Lecture Notes in Mathematics 1470, 72–22. Springer, Berlin Heidelberg Newyork (1991)
Nemirovski, A.: On tractable approximations of randomly perturbed convex constraints. In: Proceedings of the 42nd IEEE Conference on Decision and Control Maui, Hawaii USA, December 2003, pp. 2419–2422
Nemirovski A., Roos C., Terlaky T. (1999) On maximization of quadratic form over intersection of ellipsoids with common center. Math. Prog. 86, 463–473
Nemirovski, A.: Regular Banach spaces and large deviations of random sums. Working paper http://iew3.technion.ac.il/Labs/Opt/index.php?4
Shapiro A. (1985) Extremal problems on the set of nonnegative definite matrices. Linear Algebra Appl. 67, 7–18
Shapiro A., Botha J.D. (1988) Dual algorithms for orthogonal procrustes rotations. SIAM J. Matrix Anal. Appl. 9, 378–383
Ten Berge J.M.F., Nevels K. (1977) A general solution to Mosiers oblique Procrustes problem. Psychometrika 42, 593–600
Wolkowicz H. (2000) Semidefinite programming approaches to the quadratic assignment problem. In: Saigal R., Wolkowitcz H., Vandenberghe L. (eds) Handbook on Semidefinite Programming. Kluwer, Dordrecht
Wolkowicz H., Zhao Q. (1999) Semidefinite programming relaxations for graph partitioning problems. Discrete Appl. Math. 96/97: 461–479
Zhao Q., Karisch S.E., Rendl F., Wolkowicz H. (1998) Semidefinite programming relaxations for the quadratic assignment problem. J. Comb. Optim. 2, 71–109
Author information
Authors and Affiliations
Corresponding author
Additional information
Research was partly supported by the Binational Science Foundation grant #2002038.
Rights and permissions
About this article
Cite this article
Nemirovski, A. Sums of random symmetric matrices and quadratic optimization under orthogonality constraints. Math. Program. 109, 283–317 (2007). https://doi.org/10.1007/s10107-006-0033-0
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10107-006-0033-0
Keywords
- Large deviations
- Random perturbations of linear matrix inequalities
- Semidefinite relaxations
- Orthogonality constraints
- Procrustes problem