Tukey’s M-estimator of the Poisson parameter with a special focus on small means
- 244 Downloads
We treat robust M-estimators for independent and identically distributed Poisson data. We introduce modified Tukey M-estimators with bias correction and compare them to M-estimators based on the Huber function as well as to weighted likelihood and other estimators by simulation in case of clean data and data with outliers. In particular, we investigate the problem of combining robustness and high efficiencies at small Poisson means caused by the strong asymmetry of such Poisson distributions and propose a further estimator based on adaptive trimming. The advantages of the constructed estimators are illustrated by an application to smoothing count data with a time varying mean and level shifts.
KeywordsCount data Poisson model Huber M-estimator Tukey M-estimator Outliers Robustness
Mathematics Subject Classification (2000)62-07 62F35 62G05 62G35
The FCS data are kindly provided by the Department of Systemic Cell Biology of the Max Planck Institute of Molecular Physiology in Dortmund, Germany. The authors are very grateful for the stimulating comments of two reviewers and the editor which were very helpful for improving the paper.
- Agostinelli C (2013) Wle. Version 0.9-9. Available from http://cran.r-project.org/
- Cadigan NG, Chen J (2001) Properties of robust M-estimators for Poisson and negative binomial data. J Stat Comput Simul 70:273–288Google Scholar
- Chu CK, Glad IK, Godtliebsen F, Marron JS (1998) Edge-preserving smoothers for image processing. J Am Stat Assoc 93:526–541Google Scholar
- Huber P (1981) Robust statistic of location parameter. Wiley, New YorkGoogle Scholar
- Kohl M, Ruckdeschel P (2013) ROptEst. Version 0.9. Available from http://cran.r-project.org/
- R version 3.0.2 (2013) Copyright (C) 2013 The R Foundation for Statistical ComputingGoogle Scholar