Abstract
Optimization problems with orthogonality constraints are classical nonconvex nonlinear problems and have been widely applied in science and engineering. In order to solve this problem, we come up with an adaptive scaled gradient projection method. The method combines a scaling matrix that depends on the step size with some parameters that control the search direction. In addition, we consider the BB step size and combine a nonmonotone line search technique to accelerate the convergence speed of the proposed algorithm. Under the premise of non-monotonic, we prove the convergence of the algorithm. Also, the computation results proved the efficiency of the proposed algorithm.
Similar content being viewed by others
Data Availability
Not Applicable.
References
Stiefel, E.: Richtungsfelder und Fernparallelismus in n-dimensionalen Mannigfaltigkeiten. Commentarii Mathe. Helvetici 8(1), 305–353 (1935)
Elden, L., Park, H.: A Procrustes problem on the Stiefel manifold. Numer. Math. 82(4), 599–619 (1999)
Schonemann, P.H.: A generalized solution of the orthogonal Procrustes problem. Psychometrika 31(1), 1–10 (1966)
Saad, Y.: Numerical Methods for Large Eigenvalue Problems. Manchester University Press, Manchester (1992)
Wen, Z., Yang, C., Liu, X., et al.: Trace-penalty minimization for large-scale eigenspace computation. J. Sci. Comput. 66(3), 1175–1203 (2016)
Kokiopoulou, E., Chen, J., Saad, Y.: Trace optimization and eigenproblems in dimension reduction methods. Numer. Linear. Algeb. Appl. 18(3), 565–602 (2011)
Zhaosong, L., Zhang, Y.: An augmented Lagrangian approach for sparse principal component analysis. Math. Program. 135(1–2), 149–193 (2012)
Zou, H., Hastie, T., Tibshirani, R.: Sparse principal component analysis. J. Comput. Graph. Stat. 15(2), 265–286 (2006)
Theis, F.J., Cason, T.P., Absil, P.A.: Soft dimension reduction for ICA by joint diagonalization on the Stiefel manifold. Indep. Compon. Anal. Sign. Sep. 5441, 354–361 (2009)
Manton, J.H.: Optimization algorithms exploiting unitary constraints. IEEE Trans. Sign. Process. 50(3), 635–650 (2002)
Nishimori, Y., Akaho, S.: Learning algorithms utilizing quasi-geodesic flows on the Stiefel manifold. Neurocomputing 67, 106–135 (2005)
Abrudan, T.E., Eriksson, J., Koivunen, V.: Steepest descent algorithms for optimization under unitary matrix constraint. IEEE Trans. Sign. Process. 56(3), 1134–1147 (2008)
Edelman, A., Arias, T.A., Smith, S.T.: The geometry of algorithms with orthogonality constraints. SIAM J. Matrix Anal. Appl. 20(2), 303–353 (1998)
Abrudan, T., Eriksson, J., Koivunen, V.: Conjugate gradient algorithm for optimization under unitary matrix constraint. Sign. Process. 89(9), 1704–1714 (2009)
Absil, P.A., Baker, C.G., Gallivan, K.A.: Trust-region methods on Riemannian manifolds. Found. Comput. Math. 7(3), 303–330 (2007)
Savas, B., Lim, L.H.: Quasi-Newton methods on Grassmannians and multilinear approximations of tensors. SIAM J. Sci. Comput. 32(6), 3352–3393 (2010)
Absil, P.A., Mahony, R., Sepulchre, R.: Optimization Algorithms on Matrix Manifolds. Princeton University Press (2009)
Absil, P.A., Malick, J.: Projection-like retractions on matrix manifolds. SIAM J. Optim. 22(1), 135–158 (2012)
Gao, B., Liu, X., Chen, X., et al.: A new first-order framework for orthogonal constrained optimization problems. Optimization 28, 302–332 (2016)
Zhu, X.J.: A Riemannian conjugate gradient method for optimization on the Stiefel manifold. Comput. Optim. Appl. 67(1), 73–110 (2017)
Oviedo, H., Dalmau, O., Lara, H.: Two adaptive scaled gradient projection methods for Stiefel manifold constrained optimization. Num. Algorithm. 87, 1107–1127 (2020)
Wen, Z., Yin, W.: A feasible method for optimization with orthogonality constraints. Math. Program. 142(1–2), 397–434 (2013)
Chamberlain, R.M., Powell, M.J.D.: The watchdog technique for forcing conver-gence in algorithm for constrained optimization. Math. Program. Stud. 16, 1–17 (1982)
Grippo, L., Lampariello, F., Lucidi, S.: A nonmonotone line search technique for Newton’s method. SIAM J. Numer. Anal. 23(4), 707–716 (1986)
Zhang, H., Hager, W.W.A.: nonmonotone line search technique and its application to unconstrained optimization. Soc. Ind. Appl. Math. 14(4), 1043–1056 (2004)
Gu, Z., Mo, J.: Incorporating nonmonotone strategies into the trust region method for unconstrained optimization. Comput. Math. Appl. 55(9), 2158–2172 (2008)
Nishimori, Y., Akaho, S.: Learning algorithms utilizing quasi-geodesic flows on the Stiefel manifold. Neurocomputing 67(none), 106–135 (2005)
Oviedo, H., Lara, H., Dalmau, O.: A non-monotone linear search algorithm with mixed direction on Stiefel manifold. Optim. Methods Softw. 34(2), 437–457 (2019)
Boufounos, J., Borwein, J.M.: Two-point step size gradient methods. IMA J. Numer. Anal. 8(1), 141–148 (1988)
Hu, J., Liu, X., Wen, Z., Yuan, Y.: A brief introduction to manifold optimization. J. Oper. Res. Soc. China 8, 199–248 (2020)
Zhu, X.: A Riemannian conjugate gradient method for optimization on the Stiefel manifold. Comput. Optim. Appl. 67, 73–110 (2017)
Oviedo, H., Dalmau, O.: A scaled gradient projection method for minimization over the Stiefel manifold. In: International Conference on Artificial Intelligence, pp. 239–250, Springer, Cham (2019).
Sun, Y., Huang, Y.: A Riemannian conjugate gradient method on stiefel manifold. 44(3), 255–266 (2022)
Acknowledgements
The authors would be grateful for the comments and suggestions proposed by anonymous reviewers.
Funding
This work was funded by National Science Foundation of China (12171042).
Author information
Authors and Affiliations
Contributions
Q wrote and debug the code for the algorithm. The main idea of this work is put forward by Q. Both authors wrote and revised the manuscript together.
Corresponding author
Ethics declarations
Conflict of interest
The author's declared that they have no conflict of interest.
Ethical Approval
Not Applicable.
Additional information
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
Ji, Q., Zhou, Q. A Non-monotone Adaptive Scaled Gradient Projection Method for Orthogonality Constrained Problems. Int. J. Appl. Comput. Math 10, 89 (2024). https://doi.org/10.1007/s40819-024-01689-6
Accepted:
Published:
DOI: https://doi.org/10.1007/s40819-024-01689-6