Article ID Journal Published Year Pages File Type
1146898 Journal of Multivariate Analysis 2011 14 Pages PDF
Abstract

We consider the robust estimation of regression parameters in linear models with long memory and heavy-tailed errors. Asymptotic Bahadur-type representations of robust estimates are developed and their limiting distributions are obtained. It is shown that the limiting distributions are very different from those obtained under short memory. A simulation study is carried out to compare the performance of various asymptotic representations.

Related Topics
Physical Sciences and Engineering Mathematics Numerical Analysis
Authors
, ,