Asymptotic normality of p-norm estimators in multiple regression |
| |
Authors: | Arjen E Ronner |
| |
Institution: | (1) Econometric Institute, University of Groningen, P.O. Box 800, NL-9700 Groningen, The Netherlands;(2) Present address: Philips, Centre for Quantitative Methods, Eindhoven |
| |
Abstract: | Summary The consistency and asymptotic normality of p-norm estimators (1<p<2) is established by applying some of the ideas of Huber (1973), where asymptotic normality of the so-called M-estimators for regression parameters is shown. A central role is played by a weight function . Huber assumed that , ![psgr](/content/hh487x4867161128/xxlarge968.gif) and ![psgr](/content/hh487x4867161128/xxlarge968.gif) are bounded. This is, however, not the case for p-norm estimators with 1<p<2, but some of his ideas can still be applied. |
| |
Keywords: | |
本文献已被 SpringerLink 等数据库收录! |
|