Minimum Distance Estimation in AR(1)-processes

被引：0
|
作者：
Michel Piot
机构：
[1] University of Bern,Department of Mathematical Statistics
来源：
Methodology And Computing In Applied Probability | 2002年 / 4卷 / 2期
关键词：
autoregressive processes; minimum distance estimation; contaminated normal distribution; asymptotic behavior;
D O I：
10.1023/A:1020620222477
中图分类号：
学科分类号：
摘要：
In this paper a new minimum distance estimator is defined in case that the residuals of an AR(1)-process are contaminated normally distributed. This estimator is asymtotically normally distributed and in most cases less biased than the least square estimator. Furthermore, a method is presented to numerically calculate the minimum distance estimator as a root of an implicit function.
引用
收藏
页码：123 / 141
页数：18
相关论文
共 50 条
12345