Abstract
This paper studies convergence properties of regularized Newton methods for minimizing a convex function whose Hessian matrix may be singular everywhere. We show that if the objective function is LC2, then the methods possess local quadratic convergence under a local error bound condition without the requirement of isolated nonsingular solutions. By using a backtracking line search, we globalize an inexact regularized Newton method. We show that the unit stepsize is accepted eventually. Limited numerical experiments are presented, which show the practical advantage of the method.
Similar content being viewed by others
References
H. Dan, N. Yamashita, and M. Fukushima, “Convergence properties of the inexact Levenberg-Marquardt method under local error bound conditions,” Optimization Methods and Software, vol. 17, pp. 605-626, 2002.
R.S. Dembo, S.C. Eisenstat, and T. Steihaug, “Inexact Newton method,” SIAM Journal on Numerical Analysis, vol. 19, pp. 400-408, 1982.
J.Y. Fan and Y.X. Yuan, “On the convergence of a new Levenberg-Marquardt method,” Technical Report, AMSS, Chinese Academy of Sciences, 2001.
G.W. Stewart and J.G. Sun, Matrix Perturbation Theory, Academic Press: New York, NY, 1990.
N. Yamashita and M. Fukushima, “On the rate of convergence of the Levenberg-Marquardt method,” Computing, vol. 15(Suppl), pp. 239-249, 2001.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Li, DH., Fukushima, M., Qi, L. et al. Regularized Newton Methods for Convex Minimization Problems with Singular Solutions. Computational Optimization and Applications 28, 131–147 (2004). https://doi.org/10.1023/B:COAP.0000026881.96694.32
Issue Date:
DOI: https://doi.org/10.1023/B:COAP.0000026881.96694.32