A Nonparametric Method for Ascertaining Change Points in Regression Regimes

Of interest is the specific model called the joinpoint two regime regression or broken line model composed of one regression line and a horizontal ray. This is a very restricted but highly useful subset of the well-researched change point problem. The usual approach to a more general model was first presented by Quandt (1958) who found the maximum likelihood estimates of the slope, intercept and joinpoint by assuming that the error terms are generated under the usual assumptions, that is, from a normal distribution with constant variance and are uncorrelated. We develop a method that does not rely on this assumption, demonstrate its use on an example of proximity indexes of whale cow and calf pairs, and compare the new method to the Quandt estimates in a simulation study showing this new method performs adequately.

Name thesis:degree_name

Master of Science in Mathematics (M.S.)

Level thesis:degree_level

Thesis (open access)

Discipline thesis:degree_discipline

Department of Mathematical Sciences

Year dc:date.available

2010

Author dc:creator

• Rogers, Alfreda N.

Contributors dc:contributor

• Martha Abell
• Greg Knofczynski

• ETD
• Maximum likelihood
• Change point
• Moment match

Repository record dc:identifier

https://digitalcommons.georgiasouthern.edu/etd/664

OAI identifier oai:identifier

oai:digitalcommons.georgiasouthern.edu:etd-1664