|
Activity Number:
|
121
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
IMS
|
| Abstract - #304618 |
|
Title:
|
An Almost Nonparametric Model for Missing Covariates in Parametric Regression
|
|
Author(s):
|
Byungtae Seo*+
|
|
Companies:
|
Texas Tech University
|
|
Address:
|
Department of Mathematics and Statistics, Lubbock, TX, 79409-1042,
|
|
Keywords:
|
missing covariates ; semiparametric mixture ; misspecification
|
|
Abstract:
|
In missing covariate problems in parametric regression model, the ML method requires a model for the covariate distribution as well as a parametric regression model under MAR. However, the ML method is often suffered from a misspecified covariate distribution. An ideal way to circumvent this problem would be to leave the covariate distribution unspecified. We discuss this nonparametric method and investigate its potential problem. Based on understanding the inability of the ML method in such nonparametric models, we propose a new method which uses a nonparametric model for observed covariates and an almost nonparametric model for unobserved covariates.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2009 program |
