Abstract
A class of dampled Morozov principles is introduced to determine the regularization parameter for the least squares formulation of nonlinear illposed inverse problems. Their asymptotic behavior as the error level in the data converges to zero is studied and a numerical examples is given which shows that a damped Morozov principlecan be superior to the Morozov principle.
Zusammenfassung
Eine Klasse gedämpfter Morozov'scher Prinzipien zur Bestimmung des Regularisierungsparameters in schlechtgestellten nichtlinearen inversen Problemen wird eingeführt und ihr asymptotisches Verhalten wird analysiert. an Hand eines Beispiels wird gezeigt, daß das gedämpfte Morozov'sche Prinzip bessere numerische Resultate als das klassische Morozov Prinzip liefern kann.
Similar content being viewed by others
References
Banks, H. T., Kunisch, K.: Estimation techniques for distributed parameter systems. Boston: Birkhäuser 1989.
Baumeister, J.: Stable solutions of inverse problems. Braunschiweig: Vierweg 1986.
Colonius, F., Kunisch, K.: Output least squares stability in elliptic systems. Appl. Math. Opt.19, 33–63 (1989).
Engl, H. W., Kunisch, K., Neubauer, A.: Convergence rates for Tikhonov regularization of nonlinear ill-posed problems. Inverse Problems5, 523–540 (1989).
Groetsch, C. W.: The Theory of Tikhonov regularization for Fredholm equations for the first Kind. Boston: Pitman 1983.
Ito, K., Kunsich, K.: On the coefficient-to-solution mapping for elliptic boundary value problems and its linearization, to appear in J. Math. Anal. Appl.
Kunisch, K., Geymayer, G.: Convergence rates for regularized nonlinear illposed problems. To appear in Proceedings of Conference on “Modelling and Inverse Problems of Control for Distributed Parameter Systems” IIASA, Laxenburg, 1989.
Kravaris, C., Seinfeld, J.: Identification of parameters in distributed systems by regularization, SIAM J. Control and Optimization23, 271–241 (1985).
Louis, A. K.: Inverse und schlechtgestellte Probleme. Stuttgart: Teubner 1989.
Morozov, V. A.: Methods for solving incorrectly posed problems. Berlin, Heidelberg, New York, Tokyo: Springer 1984.
Neubauer, A.: Tihonov regularization for non-linear ill-posed problems: optimal convergence rates and finite-dimensional approximation. Inverse problems5, 541–558 (1989).
Neubauer, A., Scherzer, O.: Finite-dimensional approximations of Tikhonov regularized solutions of nonlinear ill-posed problems. Num. Funct. Anal. Opt. (to appear).
Author information
Authors and Affiliations
Additional information
Supported in part by the Fonds zu Förderung der wissienschiaftlichen Forschung, Austria, under P7869-PHY and by the Bundesministerium für Wissenschaft und Forschung.
Rights and permissions
About this article
Cite this article
Kunisch, K. On a class of damped Morozov principles. Computing 50, 185–198 (1993). https://doi.org/10.1007/BF02243810
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF02243810