SEMINARS
|
|
Fall 2007
STATISTICS
COLLOQUIUM
Wednesday, September 26, 2007
3:30-4:00—Refreshments
4:00-5:00—Talk
Yost Hall, Room 101
Akim Adekpedjou, PhD
Department of Mathematics and Statistics, University of Missouri-Rolla
Rolla, MO
Estimation with Recurrent Event Data: Modeling and Analysis
We consider a study [biomedical, engineering, reliability, economics, actuarial science etc…] which monitors the occurrences of a recurrent event for n subjects or units over a random observation period for each subject. We assume that the distribution of the random observation window is informative regarding the distribution of event time. The problem of parametric and semi parametric estimation of the cumulative hazard and consequently of the gap-time [inter-event time] is considered under a model of informative censoring called the Koziol-Green model. A Nelson-Aalen and Kaplan-Meier type estimators respectively for the cumulative hazard and the gap-time distribution function under the specified model are proposed. Asymptotic and small sample properties of the proposed estimators are established and compared to existing ones to ascertain efficiency gained. Different aspects of modeling and analysis as well as major areas of application of recurrent events data will be discussed.
Key words: Martingales, Covariation process, Counting processes, Stochastic integration, Efficiency
|
