Date of Award
6-1997
Degree Name
Doctor of Philosophy
Department
Mathematics
First Advisor
Dr. Joseph W . McKean
Abstract
A weighted rank-based estimate for estimating the parameter of an auto-regressive time series is considered. When the weights are constant, the estimate is equivalent to using Jaeckel’s estimate and Wilcoxon scores. The estimate can be shown to be asymptotically normal at rate y/n. In a linear regression setting this estimate has the desired properties of a continuous totally bounded influence function and a positive breakdown point. It is shown via examples and Monte Carlo that these properties are preserved in an autoregressive time series setting.
Access Setting
Dissertation-Open Access
Recommended Citation
Terpstra, Jeffrey, "A Robust Estimate for an Autoregressive Time Series" (1997). Dissertations. 1655.
https://scholarworks.wmich.edu/dissertations/1655