|
Activity Number:
|
447
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 6, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Nonparametric Statistics
|
| Abstract - #302557 |
|
Title:
|
Smooth Regression in Model Selection-Shrinkage and Average
|
|
Author(s):
|
Jingyi Liu*+
|
|
Companies:
|
University of California, Davis
|
|
Address:
|
4400 Solano Park Cir. Apt 4123, Davis, CA, 95616,
|
|
Keywords:
|
Shrinkage ; Average ; Model fitting
|
|
Abstract:
|
When a regression problem contains many predictor variables, a regression equation based on a few variables will be more accurate and certainly simpler. Shrinkage methods are widely used in the literature when most information is compressed into the first N dimensions. However, if the regression coefficients show very little difference from each other, it is not easy to choose a few dimensions to represent the data information, especially when the number of predictors has the same order as the number of observations. We develop a method called smooth regression, which includes both shrinkage and average procedure. Simulation results show that smooth regression always outperforms shrinkage alone in terms of Mean Square Error (MSE). Asymptotic approximation of the MSE of the proposed method is derived with certain assumptions of shrinkage and average functions.
|
- 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 2008 program |
