Suet Mui Tang

Date of Award


Degree Type


Degree Name

Doctor of Philosophy


The objective of this thesis is to develop methodology for detecting parameter changes at unknown times in regression models with autocorrelated errors. This methodology has many applications including quality control, environmental monitoring and medical research.;Limit processes of partial sums of residuals from stationary time series are first derived, and these results are applied to obtain limit processes of partial sums of regression residuals with stationary error structure. A class of statistics based on these partial sums is proposed for detection of time series interventions occurring at unknown times. It is shown that the large sample effects of autocorrelation on change detection statistics for regression can be accounted for by adjustment to the white noise case. The change detection statistics for regression with white noise error structures are applied to the cases with stationary processes. The asymptotic distributions of change detection statistics for AR(1) model is shown to be that of Cramer-von Mises type functionals on Brownian Bridge type processes. Selected quantiles are tabulated for a variety of these statistics.;The methodology is then applied to a hydrology example. Areas for further development are discussed.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.