Abstract
In this paper, we propose a new family of robust regression estimators, which we call bounded residual scale estimators (BRS-estimators) which are simultaneously highly robust and highly efficient for small samples with normally distributed errors. To define these estimators it is required to have a robust M-scale and a family of robust MM-estimators. We start by choosing in this family a highly robust initial estimator but not necessarily highly efficient. Loosely speaking, the BRS-estimator is defined as the estimator in the MM family which is closest to the LSE among those with a robust M-scale sufficiently close to the one of the initial estimators. The efficiency of the BRS is derived from the fact that when there are not outliers in the sample and the errors are normally distributed, the scale of the LSE is similar to the one of the initial estimator. The robustness of the BRS-estimator comes from the fact that its robust scale is close to the one of the initial highly robust estimator. The results of a Monte Carlo study show that the proposed estimator has a high finite-sample efficiency, and is highly resistant to outlier contamination.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Bondell, H.D., Stefanski, L.A.: Efficient robust regression via two-stage generalized empirical likelihood. J. Am. Stat. Assoc. 108, 644–655 (2013)
Davies, P.L.: Asymptotic behaviour of S-estimates of multivariate location parameters and dispersion matrices. Ann. Stat. 15, 1269–1292 (1987)
Davies, L.: The asymptotics of S-estimators in the linear regression model. Ann. Stat. 18, 1651–1675 (1990)
Donoho, D.L., Huber, P.J.: The notion of breakdown point. In: Festschrift for Erich L. Lehmann, pp. 157–184. Wadsworth, Belmont (1983)
Fasano M.V., Maronna R.A., Sued R.M., Yohai V.J.: Continuity and differentiability of regression M functionals. Bernoulli 18, 1289–1309 (2012)
Gervini, D., Yohai, V.J.: A class of robust and fully efficient regression estimators. Ann. Stat. 30, 583–616 (2002)
Hampel, F.R.: Beyond location parameters: robust concepts and method (with discussion). Bull. Int. Stat. Inst. 46, 375–391 (1975)
Hössjer, O.: On the optimality of S-estimators. Stat. Probab. Lett. 14, 413–419 (1992)
Huber, P.J.: Robust Statistics. Wiley, New York (1981)
Lopuhaä H.P.: Highly efficient estimators of multivariate location with high breakdown point. Ann. Stat. 20, 398–413 (1992)
Maronna, R.A., Yohai, V.J.: High finite-sample efficiency and robustness based on distance-constrained maximum likelihood. Comput. Stat. Data Anal. (2014). doi:10.1016/j.csda.2014.10.015
Maronna, R.A., Martin, R.D., Yohai, V.J.: Robust Statistics: Theory and Methods. Wiley, New York (2006)
Rousseeuw, P.J.: Least median of squares regression. J. Am. Stat. Assoc. 79, 871–880 (1984)
Rousseeuw, P.J., Leroy, A.: Robust Regression and Outlier Detection. Wiley, New York (1987)
Rousseeuw, P.J., Yohai, V.J.: Robust Regression by Means of S-Estimators. Lecture Notes in Statistics, vol. 26, pp. 256–272. Springer, New York (1984)
Stahel, W.A., Yohai, V.J., Zamar, R.H.: A procedure for robust estimation and inference in linear regression. In: Stahel, W., Weisberg, S. (eds) Directions in Robust Statistics and Diagnostics (Part II). The IMA Volumes in Mathematics and Its Applications. Springer, New York (1991)
Yohai, V.J.: High Breakdown point and high efficiency estimates for regression. Ann. Stat. 15, 642–656 (1987)
Yohai, V.J., Zamar, R.H.: High breakdown point estimates of regression by means of an efficient scale. Technical Report No. 84, Department of Statistics, University of Washington, Seattle. Available at https://www.stat.washington.edu/research/reports/1986/tr084.pdf (1986)
Acknowledgements
This research was partially supported by Grants W276 from Universidad of Buenos Aires, PIPs 112-2008-01-00216 and 112-2011-01- 00339 from CONICET and PICT 2011-0397 from ANPCYT, Argentina. Ezequiel Smucler was supported by a doctoral scholarship of the University of Buenos Aires. We are grateful to two anonymous referees, whose comments lead to improvements in the paper.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2015 Springer International Publishing Switzerland
About this chapter
Cite this chapter
Smucler, E., Yohai, V.J. (2015). Highly Robust and Highly Finite Sample Efficient Estimators for the Linear Model. In: Nordhausen, K., Taskinen, S. (eds) Modern Nonparametric, Robust and Multivariate Methods. Springer, Cham. https://doi.org/10.1007/978-3-319-22404-6_6
Download citation
DOI: https://doi.org/10.1007/978-3-319-22404-6_6
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-22403-9
Online ISBN: 978-3-319-22404-6
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)