Models of changes in autoregressive sequences

Abstract

This thesis deals with the detection of change in the structure of an autoregressive time series. In the first part of the thesis we provide an overview of the main results concerning the theory of autoregressive processes (Chapter 1) and a general theory of the maximum likelihood approach towards the change point problem (Chapter 2). The second and main part of the thesis (Chapter 3) deals with various approaches to the CPP applied on the autoregressive processes and provides a comprehensive proof of a theorem that was previously published by Hušková et al. (2007a) only with a sketched proof. A third part of the thesis contains a computer simulation study of the performance of the studied statistics (Chapter 4). Two appendices contain most of our proofs and also some general results of probability theory and statistics that were used in the thesis.