|
Activity Number:
|
60
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Sunday, July 29, 2007 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #310411 |
|
Title:
|
Improved Centroids Estimation for the Nearest Shrunken Centroid Classifier
|
|
Author(s):
|
Sijian Wang*+ and Ji Zhu
|
|
Companies:
|
University of Michigan and University of Michigan
|
|
Address:
|
1420 Washington Heights, Ann Arbor, MI, 48109-2029,
|
|
Keywords:
|
High Dimension Low Sample Size ; LASSO ; Microarray ; Nearest Shrunken Centroid ; Regularization ; Variable Selection
|
|
Abstract:
|
The nearest shrunken centroid (NSC) method has been successfully applied in many DNA microarray classification problems. We show that the NSC method can be interpreted in the framework of LASSO regression. Based on that, we propose two new methods (or penalty functions) for microarray classification, which improve over the NSC. Unlike the L1-norm penalty used in LASSO, the penalty terms that we propose make use of the fact that parameters belonging to one gene should be treated as a natural group. Numerical results indicate that the two new methods tend to remove irrelevant variables more effectively and provide better classification results than the L1-norm approach.
|
