|
Activity Number:
|
153
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 7, 2006 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #307094 |
|
Title:
|
A Partially Exchangeable Model and Its Applications in Correlated Data
|
|
Author(s):
|
Latonya Garner*+ and Hanxiang Peng
|
|
Companies:
|
University of Mississippi and University of Mississippi
|
|
Address:
|
P.O. Box 3958, University, MS, 38677,
|
|
Keywords:
|
|
|
Abstract:
|
In this talk, we generalize the binomial model by relaxing independence to partial exchangeability. We derive the distribution when the sequence of Bernoulli trials forms a Markov chain. We show that the generalized binomial distribution by George and Bowman (1995) hence the classical binomial distribution is a special case. A simulation study will be conducted to illustrate that the proposed distribution can be used to study correlated data, which is our motivation and our future research.
|
