Abstract
A conditionally contaminated linear model y = x T ß+z is considered where the errors z may have different contaminated normal distributions for different experimental conditions x. At this model one-step M-estimators have an asymptotic bias which should be bounded for robust estimators. For designs which have a support of linearly independent regressors, an algorithm for the calculation of one-step M-estimators which minimize the asymptotic mean squared error (MSE) is presented. The results of this algorithm are given for quadratic regression and for a problem in a one-way lay-out model.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
BICKEL, P.J. (1975). One-step Huber estimates in the linear model. J. Amer. Statist. Assoc. 70, 428–434.
BICKEL, P.J. (1984). Robust regression based on infinitesimal neighbourhoods. Ann. Statist. 12, 1349–1368.
HAMPEL, F.R. (1978). Optimally bounding the gross-error-sensitivity and the influence of position in factor space. Proceedings of the ASA Statistical Computing Section, ASA, Washington, D.C., 59–64.
KRASKER, W.S. (1980). Estimation in linear regression models with disparate data points. Econometrica 48, 1333–1346.
KUROTSCHKA, V. and MÜLLER, Ch.H. (1992). Optimum robust estimation of linear aspects in conditionally contaminated linear models. Ann. Statist. 20, 331–350.
MÜLLER, Ch.H. (1987). Optimale Versuchspläne für robuste Schätzfunktionen in linearen Modellen. Ph. D. thesis. Freie Universität Berlin.
MÜLLER, Ch.H. (1992). One-step-M-estimators in conditionally contaminated linear models. Preprint No. A-92–11, Freie Universität Berlin, Fachbereich Mathematik. Submitted to Stat. Decis.
MÜLLER, Ch.H. (1993a). Behaviour of asymptotically optimal designs for robust estimation at finite sample sizes. In: Model-Oriented Data Analysis, eds. W.G. Müller, H.P. Wynn, A.A. Zhigljaysky, 53–62, Physica-Verlag, Heidelberg.
MÜLLER, Ch.H. (1993b). Optimal bias bounds for robust estimation in linear models. Preprint No. A-35–93, Freie Universität Berlin, Fachbereich Mathematik und Informatik. To appear in the Proceedings of the LINSTAT’93 Conference, Poznan.
MÜLLER, Ch.II. (1994). Optimal designs for robust estimation in conditionally contaminated linear models. J. Statist. Plann. Inference. 38, 125–140.
RIEDER, H. (1987). Robust regression estimators and their least favorable contamination curves. Stat. Decis. 5, 307–336.
RIEDER, H. (1993). Robust Asymptotic Statistics, Springer, New York, to appear.
SAMAROV, A.M. (1985). Bounded-influence regression via local minimax mean squared error. J. Amer. Statist. Assoc. 80, 1032–1040.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1994 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Müller, C.H. (1994). On the Calculation of MSE Minimizing Robust Estimators. In: Dutter, R., Grossmann, W. (eds) Compstat. Physica, Heidelberg. https://doi.org/10.1007/978-3-642-52463-9_29
Download citation
DOI: https://doi.org/10.1007/978-3-642-52463-9_29
Publisher Name: Physica, Heidelberg
Print ISBN: 978-3-7908-0793-6
Online ISBN: 978-3-642-52463-9
eBook Packages: Springer Book Archive