Dissertation
Change-point analysis: Single change-point in a sequence of independent Gaussian and exponential random variables
Washington State University
Doctor of Philosophy (PhD), Washington State University
05/2009
DOI:
https://doi.org/10.7273/000006181
Abstract
This thesis considers the problem of change-point analysis, more specifically the maximum likelihood estimates of a single abrupt change-point in sequence of independent, time-ordered exponential and normal random variables and the asymptotic distributions of those estimates. Exact computable expressions for the asymptotic distributions are calculated and the distributions are used to calculate confidence intervals for a change detected in the multivariate gaussian time-series of annual mean precipitation, a univariate gaussian time-series of temperature anomalies, and a time-ordered sequence of exponentially distributed time intervals between earthquakes. The accuracy of the asymptotic distribution and, in the gaussian case, robustness to departures from normality are investigated via simulations.
Metrics
1 File views/ downloads
12 Record Views
Details
- Title
- Change-point analysis
- Creators
- Elena A. Khapalova
- Contributors
- Stergios B. Fotopoulos (Chair)Venkata Krishna Jandhyala (Committee Member) - Washington State University, Department of Mathematics and StatisticsSung Keuk Ahn (Committee Member) - Washington State University, Department of Finance and Management ScienceMin-Chiang Wang (Committee Member) - Washington State University, Conversion (Inactive)Kuruppu A Ariyawansa (Committee Member) - Washington State University, Department of Mathematics and Statistics
- Awarding Institution
- Washington State University
- Academic Unit
- Carson College of Business
- Theses and Dissertations
- Doctor of Philosophy (PhD), Washington State University
- Publisher
- Washington State University
- Number of pages
- 205
- Identifiers
- 99901055121201842
- Language
- English
- Resource Type
- Dissertation