|
Activity Number:
|
226
|
|
Type:
|
Invited
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
IMS
|
| Abstract - #307919 |
|
Title:
|
Selecting Models Using Incomplete Data
|
|
Author(s):
|
Gerda Claeskens*+
|
|
Companies:
|
K.U. Leuven
|
|
Address:
|
ORSTAT, Leuven, International, B-3000, Belgium
|
|
Keywords:
|
model selection ; missing data ; AIC
|
|
Abstract:
|
We propose model selection methods which can deal with missing or incomplete data. More particularly, we focus on sets of data where the response variable is always observed, but some of the covariates are possibly missing. The relevant model selection question is which of the covariates should be included in the final model. We study the use of the EM algorithm, employing a method of weights, in combination with model selectors in the spirit of Akaike's information criterion AIC. In a first step the missing data mechanism is assumed to be ignorable. Extensions to other missingness mechanisms are possible. The model selection method is tested in a simulation study and illustrated by data analysis.
|
- 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 2007 program |
