|
Activity Number:
|
124
|
|
Type:
|
Invited
|
|
Date/Time:
|
Monday, July 30, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #308112 |
|
Title:
|
Simultaneous Regression Shrinkage, Variable Selection, and Supervised Clustering
|
|
Author(s):
|
Howard D. Bondell*+
|
|
Companies:
|
North Carolina State University
|
|
Address:
|
Department of Statistics, Raleigh, NC, 27695,
|
|
Keywords:
|
correlation ; penalization ; regression ; shrinkage ; supervised clustering ; variable selection
|
|
Abstract:
|
A new penalization technique is proposed to simultaneously select variables and perform supervised clustering. The form of the penalty function accomplishes variable selection by shrinking some coefficients to exactly zero. Additionally, this penalty yields exact equality of some coefficients, encouraging correlated predictors with similar effects on the response to form predictive clusters represented by a single coefficient. Pre-specification of the predictive clusters is not needed, thus performing the supervised clustering task within the estimation. This penalized likelihood estimator can also be viewed as the posterior mode for a particular choice of prior distribution. The procedure can be used in both regression and classification problems and compares favorably with existing approaches while yielding the added grouping information not given by typical procedures.
|
- 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 |
