|
Activity Number:
|
17
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Sunday, July 29, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
IMS
|
| Abstract - #308606 |
|
Title:
|
Principal Components for Regression: A Conditional Point of View
|
|
Author(s):
|
Liliana Forzani*+
|
|
Companies:
|
The University of Minnesota
|
|
Address:
|
313 Ford Hall, School of Statistics, Minneapolis, MN, 55455,
|
|
Keywords:
|
Dimension Reduction ; Principal component analysis
|
|
Abstract:
|
Many statistical applications involve regression with many predictors and problems arise when the number of predictors is large or they are very correlated. One way to tackle the problem is to reduce the dimension of the predictors. The focus of this talk is reduction of the dimension of the predictors from the Inverse Regression point of view. Assuming that the inverse regression X|Y follows a normal distribution with covariance independent of Y, Cook (2007) was able to find the linear combinations of the predictors X that are sufficient for the regression of X, in the sense that the distribution of Y|X is the same than the distribution of Y|ß'X for ß e Rpxd with d< p the smallest possible number. In this talk we present the maximum likelihood estimators for such combinations as well as testing procedures for this setting.
|
- 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 |
