Abstract
Orthogonality constrained optimization is widely used in applications from science and engineering. Due to the non-convex orthogonality constraints, many numerical algorithms often can hardly achieve the global optimality. We aim at establishing an efficient scheme for finding global minimizers under one or more orthogonality constraints. The main concept is based on the noisy gradient flow constructed from stochastic differential equations (SDEs) on the Stiefel manifold, the differential geometric characterization of orthogonality constraints. We derive an explicit representation of SDE on the Stiefel manifold endowed with a canonical metric and propose a numerically efficient scheme to simulate this SDE based on Cayley transformation with a theoretical convergence guarantee. The convergence to global optimizers is proved under second-order continuity. The effectiveness and efficiency of the proposed algorithms are demonstrated on a variety of problems including homogeneous polynomial optimization, bi-quadratic optimization, stability number computation, and 3D structure determination from common lines in Cryo-EM.
Similar content being viewed by others
Notes
Note that the SDE (7) is equivalent to the one with Itô’s calculus, i.e., \({\mathrm {d}}x(t) = -\nabla f(x(t)) {\mathrm {d}}t + \sigma (t) {\mathrm {d}}B(t)\) since the diffusion coefficient does not involve the space term. The latter is often used in literature due to the popularity and the ease of formulation of Itô’s calculus. We intend to adopt the Stratonovich calculus in order to keep consistency with the manifold-constrained case discussed in subsequent sections.
For example, when we say we compute m runs of RS-local with N cycles for each, we actually generate mN random initial points and call local solver for mN times, and every N instances are grouped together to give the best solution within one run.
References
Absil, P.A., Mahony, R., Sepulchre, R.: Optimization Algorithms on Matrix Manifolds. Princeton University Press, Princeton (2008)
Aharon, M., Elad, M., Bruckstein, A.: K-SVD: an algorithm for designing overcomplete dictionaries for sparse representation. IEEE Trans. Signal Process. 54, 4311–4322 (2006)
Aluffi-Pentini, F., Parisi, V., Zirilli, F.: Global optimization and stochastic differential equations. J. Optim. Theory Appl. 47, 1–16 (1985)
Barzilai, J., Borwein, J.M.: Two-point step size gradient methods. IMA J. Numer. Anal. 8, 141–148 (1988)
Boufounos, P.T., Baraniuk, R.G.: 1-Bit compressive sensing. In: 42nd Annual Conference on Information Sciences and Systems, CISS 2008, pp. 16–21. IEEE (2008)
Cai, J.-F., Ji, H., Shen, Z., Ye, G.-B.: Data-driven tight frame construction and image denoising. Appl. Comput. Harmon. Anal. 37, 89–105 (2014)
Chiang, T.-S., Hwang, C.-R., Sheu, S.J.: Diffusion for global optimization in \(\mathbb{R}^n \). SIAM J. Control Optim. 25, 737–753 (1987)
Chow, S.-N., Yang, T.-S., Zhou, H.-M.: Global optimizations by intermittent diffusion. In: Adamatzky, A., Chen, G. (eds.) Chaos, CNN, Memristors and Beyond, pp. 466–479. World Scientific, Singapore (2013)
Edelman, A., Arias, T.A., Smith, S.T.: The geometry of algorithms with orthogonality constraint. SIAM J. Matrix Anal. Appl. 20, 303–353 (1998)
Geman, S., Hwang, C.-R.: Diffusions for global optimization. SIAM J. Control Optim. 24, 1031–1043 (1986)
Gidas, B.: Global optimization via the Langevin equation. In: 24th IEEE Conference on Decision and Control, pp. 774–778. IEEE (1985)
Goldfarb, D., Wen, Z., Yin, W.: A curvilinear search method for p-harmonic flows on spheres. SIAM J. Imaging Sci. 2, 84–109 (2009)
Gu, X., Yau, S.-T.: Global conformal surface parameterization. In: Proceedings of the 2003 Eurographics/ACM SIGGRAPH Symposium on Geometry Processing, pp. 127–137. Eurographics Association (2003)
Hairer, E., Lubich, C., Wanner, G.: Geometric Numerical Integration: Structure-Preserving Algorithms for Ordinary Differential Equations, vol. 31. Springer, Berlin (2006)
Helmberg, C., Rendl, F.: A spectral bundle method for semidefinite programming. SIAM J. Optim. 10, 673–696 (2000)
Henkel, O.: Sphere-packing bounds in the Grassmann and Stiefel manifolds. Trans. Inf. Theory 51, 3445–3456 (2005)
Hsu, E.: Stochastic Analysis on Manifolds, Vol. 38 of Graduate Studies in Mathematics, vol. 38. American Mathematical Society, Providence (2002)
Ksendal, B.Ø.: Stochastic Differential Equations, Universitext, 6th edn. Springer, Berlin (2003)
Lai, R., Wen, Z., Yin, W., Gu, X., Lui, L.M.: Folding-free global conformal mapping for genus-0 surfaces by harmonic energy minimization. J. Sci. Comput. 58, 705–725 (2014)
Laska, J.N., Wen, Z., Yin, W., Baraniuk, R.G.: Trust, but verify: fast and accurate signal recovery from 1-bit compressive measurements. IEEE Trans. Signal Process. 59, 5289–5301 (2011)
Lin, S.-Y., Luskin, M.: Relaxation methods for liquid crystal problems. SIAM J. Numer. Anal. 26, 1310–1324 (1989)
Ling, C., Nie, J., Qi, L., Ye, Y.: Biquadratic optimization over unit spheres and semidefinite programming relaxations. SIAM J. Optim. 20, 1286–1310 (2009)
Liu, X., Srivastava, A.: Stochastic search for optimal linear representations of images on spaces with orthogonality constraints. In: Rangarajan, A., Figueiredo, M., Zerubia, J. (eds.) Lecture Notes in Computer Science, pp. 3–20. Springer, Berlin (2003)
Liu, X., Srivastava, A., Gallivan, K.: Optimal linear representations of images for object recognition. In: IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Proceedings. IEEE Computer Society (2003)
Liu, X., Wen, Z., Wang, X., Ulbrich, M., Yuan, Y.: On the analysis of the discretized Kohn–Sham density functional theory. SIAM J. Numer. Anal. 53, 1758–1785 (2015)
Malham, S.J., Wiese, A.: Stochastic lie group integrators. SIAM J. Sci. Comput. 30, 597–617 (2008)
Markowich, P.A., Villani, C.: On the trend to equilibrium for the Fokker–Planck equation: an interplay between physics and functional analysis. Mat. Contemp. 19, 1–29 (2000)
Motzkin, T.S., Straus, E.G.: Maxima for graphs and a new proof of a theorem of Turán. Can. J. Math. 17, 533–540 (1965)
Munthe-Kaas, H.: Runge–Kutta methods on lie groups. BIT Numer. Math. 38, 92–111 (1998)
Nocedal, J., Wright, S.J.: Numerical Optimization, Springer Series in Operations Research and Financial Engineering, 2nd edn. Springer, New York (2006)
Oprea, J.: Differential Geometry and Its Applications, Classroom Resource Materials Series, 2nd edn. Mathematical Association of America, Washington (2007)
Ozoliņš, V., Lai, R., Caflisch, R., Osher, S.: Compressed modes for variational problems in mathematics and physics. Proc. Natl. Acad. Sci. 110, 18368–18373 (2013)
Parpas, P., Rustem, B.: An algorithm for the global optimization of a class of continuous minimax problems. J. Optim. Theory Appl. 141, 461–473 (2008)
Parpas, P., Rustem, B.: Convergence analysis of a global optimization algorithm using stochastic differential equations. J. Glob. Optim. 45, 95–110 (2009)
Parpas, P., Rustem, B., Pistikopoulos, E.N.: Linearly constrained global optimization and stochastic differential equations. J. Glob. Optim. 36, 191–217 (2006)
Parpas, P., Rustem, B., Pistikopoulos, E.N.: Global optimization of robust chance constrained problems. J. Glob. Optim. 43, 231–247 (2007)
Rubinshtein, E., Srivastava, A.: Optimal linear projections for enhancing desired data statistics. Stat. Comput. 20, 267–282 (2009)
Singer, A., Shkolnisky, Y.: Three-dimensional structure determination from common lines in Cryo-EM by eigenvectors and semidefinite programming. SIAM J. Imaging Sci. 4, 543–572 (2011)
Sloane, NJA.: Challenge problems: independent sets in graphs. https://oeis.org/A265032/a265032.html (2015). Accessed May 2019
Stratonovich, R.L.: A new representation for stochastic integrals and equations. SIAM J. Control 4, 362–371 (1966)
Tang, B., Sapiro, G., Caselles, V.: Color image enhancement via chromaticity diffusion. IEEE Trans. Image Process. 10, 701–707 (2001)
Vese, L.A., Osher, S.J.: Numerical methods for p-harmonic flows and applications to image processing. SIAM J. Numer. Anal. 40, 2085–2104 (2002)
Villani, C.: Optimal Transport, Vol. 338 of Grundlehren der mathematischen Wissenschaften, vol. 338. Springer, Berlin (2009)
Wen, Z., Milzarek, A., Ulbrich, M., Zhang, H.: Adaptive regularized self-consistent field iteration with exact Hessian for electronic structure calculation. SIAM J. Sci. Comput. 35, A1299–A1324 (2013)
Wen, Z., Yin, W.: A feasible method for optimization with orthogonality constraints. Math. Program. 142, 397–434 (2012)
Yin, G., Yin, K.: Global optimization using diffusion perturbations with large noise intensity. Acta Math. Appl. Sin. Engl. Ser. 22, 529–542 (2006)
Acknowledgements
R. Lai’s work was supported in part by NSF DMS-1522645 and an NSF Career Award DMS-1752934. Z. Wen’s work was supported in part by National Natural Science Foundation of China (Grant Nos. 11831002, 91730302 and 11421101).
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Yuan, H., Gu, X., Lai, R. et al. Global Optimization with Orthogonality Constraints via Stochastic Diffusion on Manifold. J Sci Comput 80, 1139–1170 (2019). https://doi.org/10.1007/s10915-019-00971-w
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10915-019-00971-w
Keywords
- Orthogonality constrained optimization
- Global optimization
- Stochastic differential equations
- Stochastic diffusion on manifold