Mathematics - Colloquium
Friday, November 9, 2012
11:00 AM-12:00 PM
ABSTRACT: A change in parameters of a model often characterizes occurrence of a major event like a stock market crash, breakdown of quality or a suspected major latent change. Statistical inference about the existence of such a change for a time sequence and the time when the change occurs, namely, a change-point (CP), is always the subject of considerable interest. In a classical change-point problem, the number of change-points on the data under consideration is assumed to be fixed, usually at most one change-point(AMOC), whereas, in an isotonic change-point problem, the parameters of the model are assumed to have an arbitrarily monotone structure, but the number of change-points can be as many as the number of observations in data, a typical example of which is the global warming issue. In this talk, I will introduce these two change-point problems and present my research results in these two topics.
Suggested Audiences: Adult, College
Last Modified: September 24, 2012 at 3:24 PM