Virginia Commonwealth University
Changepoint detection in correlated categorical time series
Abstract
dc:description.abstractThis dissertation develops two changepoint tests for serially correlated ordinal categorical time series. A cumulative sum type test is devised to test for a single changepoint in a correlated categorical data sequence, which is constructed from a latent Gaussian process through clipping techniques. A sequential parameter estimation procedure is proposed to estimate the model parameters. The method is illustrated via simulations and applied to a real categorized rainfall time series from Albuquerque, New Mexico. The second changepoint test is a likelihood ratio type test based on a marginalized transition model. This model permits likelihood inference and the serial dependence is modeled via a first-order Markov chain. An efficient algorithm is proposed to obtain the maximum likelihood estimates. The Lu and Wang (2012) changepoint detection method for annual frequencies of sky-cloudiness conditions at Fort St. John Airport in Canada is extended to hourly cloud cover data.
Degree
thesis:*- Name thesis:degree_name
- Doctor of Philosophy
- Level thesis:degree_level
- Dissertation
- Discipline thesis:degree_discipline
- Systems Modeling and Analysis
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- LI, MO
- Contributors dc:contributor
-
- QiQi Lu
- D'Arcy P. Mays
- Yanjun Qian
- Qin Wang
Subjects
dc:subject × 10Rights
dc:rights- Statement dc:rights
-
- © Mo Li
Identifiers
dc:identifier.*- Identifier
- https://scholarscompass.vcu.edu/etd/6810