This paper presents Riemannian conjugate gradient methods and global convergence analyses under the strong Wolfe conditions. The main idea of the proposed methods is to combine the good global convergence properties of the Dai–Yuan method with the efficient numerical performance of the Hestenes–Stiefel method. One of the proposed algorithms is a generalization to Riemannian manifolds of the hybrid conjugate gradient method of the Dai and Yuan in Euclidean space. The proposed methods are compared well numerically with the existing methods for solving several Riemannian optimization problems. Python implementations of the methods used in the numerical experiments are available at https://github.com/iiduka-researches/202008-hybrid-rcg.
This is a preview of subscription content, access via your institution.
Buy single article
Instant access to the full article PDF.
Price includes VAT (USA)
Tax calculation will be finalised during checkout.
Availability of data and materials
Absil, P.-A., Gallivan, K.A.: Joint diagonalization on the oblique manifold for independent component analysis. In: 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings, vol. 5, p. V (2006)
Absil, P.-A., Mahony, R., Sepulchre, R.: Optimization Algorithms on Matrix Manifolds. Princeton University Press, Princeton (2008)
Al-Baali, M.: Descent property and global convergence of the Fletcher–Reeves method with inexact line search. IMA J. Numer. Anal. 5(1), 121–124 (1985)
Dai, Y.-H., Yuan, Y.: A nonlinear conjugate gradient method with a strong global convergence property. SIAM J. Optim. 10(1), 177–182 (1999)
Dai, Y.-H., Yuan, Y.: An efficient hybrid conjugate gradient method for unconstrained optimization. Ann. Oper. Res. 103(1–4), 33–47 (2001)
Dolan, E.D., Moré, J.J.: Benchmarking optimization software with performance profiles. Math. Program. 91(2), 201–213 (2002)
Fletcher, R., Reeves, C.M.: Function minimization by conjugate gradients. Comput. J. 7(2), 149–154 (1964)
Hager, W.W., Zhang, H.: A survey of nonlinear conjugate gradient methods. Pac. J. Optim. 2(1), 35–58 (2006)
Hawe, S., Kleinsteuber, M., Diepold, K.: Analysis operator learning and its application to image reconstruction. IEEE Trans. Image Process. 22(6), 2138–2150 (2013)
Hestenes, M.R., Stiefel, E.: Methods of Conjugate Gradients for Solving Linear Systems. NBS, Washington (1952)
Hu, Y., Storey, C.: Global convergence result for conjugate gradient methods. J. Optim. Theory Appl. 71(2), 399–405 (1991)
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)
Polak, E., Ribière, G.: Note sur la convergence de méthodes de directions conjuguées. ESAIM Math. Model. Numer. Anal.-Modélisation Mathématique et Analyse Numérique 3(R1), 35–43 (1969)
Ring, W., Wirth, B.: Optimization methods on Riemannian manifolds and their application to shape space. SIAM J. Optim. 22(2), 596–627 (2012)
Sato, H.: A Dai–Yuan-type Riemannian conjugate gradient method with the weak Wolfe conditions. Comput. Optim. Appl. 64(1), 101–118 (2016)
Sato, H., Iwai, T.: A new, globally convergent Riemannian conjugate gradient method. Optimization 64(4), 1011–1031 (2015)
Selvan, S.E., Amato, U., Gallivan, K.A., Qi, C., Carfora, M.F., Larobina, M., Alfano, B.: Descent algorithms on oblique manifold for source-adaptive ICA contrast. IEEE Trans. Neural Netw. Learn. Syst. 23(12), 1930–1947 (2012)
Smith, S.T.: Optimization techniques on Riemannian manifolds. Fields Inst. Commun. 3(3), 113–135 (1994)
Touati-Ahmed, D., Storey, C.: Efficient hybrid conjugate gradient techniques. J. Optim. Theory Appl. 64(2), 379–397 (1990)
Townsend, J., Koep, N., Weichwald, S.: Pymanopt: a python toolbox for optimization on manifolds using automatic differentiation. J. Mach. Learn. Res. 17(1), 4755–4759 (2016)
Vandereycken, B.: Low-rank matrix completion by Riemannian optimization. SIAM J. Optim. 23(2), 1214–1236 (2013)
Wolfe, P.: Convergence conditions for ascent methods. SIAM Rev. 11(2), 226–235 (1969)
Wolfe, P.: Convergence conditions for ascent methods. II: some corrections. SIAM Rev. 13(2), 185–188 (1971)
Yuan, H., Gu, X., Lai, R., Wen, Z.: Global optimization with orthogonality constraints via stochastic diffusion on manifold. J. Sci. Comput. 80(2), 1139–1170 (2019)
We are sincerely grateful to the editor and the anonymous reviewer for helping us improve the original manuscript. This work was supported by JSPS KAKENHI Grant Number JP18K11184.
Conflict of interest
The authors declare that they have no conflict of interest.
Python implementations of the methods used in the numerical experiments are available at https://github.com/iiduka-researches/202008-hybrid-rcg.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Sakai, H., Iiduka, H. Hybrid Riemannian conjugate gradient methods with global convergence properties. Comput Optim Appl 77, 811–830 (2020). https://doi.org/10.1007/s10589-020-00224-9
- Conjugate gradient method
- Riemannian optimization
- Hybrid conjugate gradient method
- Global convergence
- Strong Wolfe conditions