Abstract
A subproblem in the trust region algorithm for non-linear M-estimation by Ekblom and Madsen is to find the restricted step. It is found by calculating the M-estimator of the linearized model, subject to anL 2-norm bound on the variables. In this paper it is shown that this subproblem can be solved by applying Hebden-iterations to the minimizer of the Lagrangian function. The new method is compared with an Augmented Lagrange implementation.
Similar content being viewed by others
References
O. Edlund, H. Ekblom, and K. Madsen,Algorithms for non-linear M-estimation, Research Report 1995-07, Department of Mathematics, Luleå University of Technology, Luleå, Sweden, 1995.
H. Ekblom and K. Madsen,Algorithms for non-linear Huber estimation, BIT, 29 (1989), pp. 60–76.
R. Fletcher,Practical Methods of Optimization, 2nd ed, John Wiley & Sons, 1987.
P. E. Gill, W. Murray, and M. H. Wright,Practical Optimization, Academic Press, 1981.
M. D. Hebden,An algorithm for minimization using exact second derivatives, Report TP515, Atomic Energy Research Establishment, Harwell, England, 1973.
K. Madsen,A combined Gauss-Newton and Quasi-Newton method for non-linear least squares, Tech. Report NI-88-10 Institute for Numerical Analysis, Technical University of Denmark, Lyngby, Denmark, 1988.
K. Madsen and H. B. Nielsen,Finite algorithms for robust linear regression, BIT, 30 (1990), pp. 333–356.
J. J. Moré,The Levenberg-Marquardt algorithm: Implementation and theory, in Numerical Analysis, Proceedings Biennial Conference Dundee 1977, G. A. Watson, ed., Berlin, 1978, Springer-Verlag, pp. 105–116.
G. A. F. Seber and C. J. Wild,Nonlinear Regression, John Wiley & Sons, 1989.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Edlund, O. Linear M-estimation with bounded variables. Bit Numer Math 37, 13–23 (1997). https://doi.org/10.1007/BF02510169
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF02510169