Some pathological regression asymptotics under stable conditions
We consider a simple through-the-origin linear regression example introduced by P. J. Rousseeuw, S. van Aelst and M. Hubert. It is shown that the conventional least squares and least absolute error estimators converge in distribution without normalization and consequently are inconsistent. A class of weighted median regression estimators, including the maximum depth estimator of Rousseeuw and Hubert [\textit{P. J. Rousseeuw} and \textit{M. Hubert}, J. Am. Stat. Assoc. 94, No. 446, 388--433 (1999; Zbl 1007.62060)], are shown to converge at rate \(n^{-1}\). Finally, the maximum likelihood estimator is considered, and we observe that there exist estimators that converge at rate \(n^{-2}\). The results illustrate some interesting, albeit somewhat pathological, aspects of stable-law convergence.
- The asymptotics of S-estimators in the linear regression model
- Consistency for least squares regression estimators with infinite variance data
- scientific article; zbMATH DE number 1215457
- Does asymptotic linearity of the regression extend to stable domains of attraction?
- scientific article; zbMATH DE number 1054299
- A Method for Simulating Stable Random Variables
- A study of logspline density estimation
- scientific article; zbMATH DE number 3828921 (Why is no real title available?)
- scientific article; zbMATH DE number 3605767 (Why is no real title available?)
- scientific article; zbMATH DE number 204994 (Why is no real title available?)
- scientific article; zbMATH DE number 3332032 (Why is no real title available?)
- On the Chambers-Mallows-Stuck method for simulating skewed stable random variables
- Parameterizations and modes of stable distributions
- Regression Depth
- Regression Quantiles
This page was built for publication: Some pathological regression asymptotics under stable conditions
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1591159)