|
Activity Number:
|
417
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #309217 |
|
Title:
|
Semiparametric Two-Sample Changepoint Model
|
|
Author(s):
|
Zonghui Hu*+ and Jing Qin and Dean Follmann
|
|
Companies:
|
National Institutes of Health and National Institutes of Health and National Institute of Allergy and Infectious Diseases
|
|
Address:
|
6700A Rockeldge Drive, Bethesda, MD, 20817,
|
|
Keywords:
|
changepoint ; empirical maximum likelihood ; likelihood ratio ; semiparametric
|
|
Abstract:
|
A semiparametric changepoint model is proposed to investigate the possible difference between two distributions. The empirical maximum likelihood (EML) method is adopted for the estimation of the changepoint and the two marginal distributions. It is shown that the EML estimator of the change-point is consistent. The likelihood ratio test for the existence of changepoint is provided, as well as the goodness-of-fit test on the changepoint model. We provide an application in a clinical study to evaluate the equivalence of two devices for measuring the viral load. Though the two devices were generally considered as equivalent for underlying viral load of large values, we showed that the equivalence fails for low viral loads.
|
