|
Activity Number:
|
419
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 9, 2006 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Nonparametric Statistics
|
| Abstract - #307302 |
|
Title:
|
Empirical Likelihood and Marginal Confidence Interval
|
|
Author(s):
|
Mi-Ok Kim*+
|
|
Companies:
|
University of Kentucky
|
|
Address:
|
817 Patterson Office Tower, Lexington, KY, 40506-0027,
|
|
Keywords:
|
empirical likelihood ; censored regression
|
|
Abstract:
|
Empirical likelihood (EL) is a nonparametric inference method with results that are in general similar to those about likelihood ratio tests and Wilk's theorem in the parametric model. In regression the EL method faces a challenge of maximizing the likelihood in presence of nuisance parameters when a single component of the regression coefficients is of interest. We propose a simple solution to the problem in the least squares regression and investigate whether the proposed method can be generalized to censored case.
|
