|
Activity Number:
|
355
|
|
Type:
|
Invited
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statisticians in Defense and National Security
|
| Abstract - #308086 |
|
Title:
|
SigClust: Statistical Significance of Clustering
|
|
Author(s):
|
J. Stephen Marron*+
|
|
Companies:
|
The University of North Carolina at Chapel Hill
|
|
Address:
|
Department of Statisitcs & O.R., Chapel Hill, NC, 27519-3260,
|
|
Keywords:
|
SigClust ; clustering ; HDLSS
|
|
Abstract:
|
Clusterings methods provide a powerful tool for the analysis of high-dimension, low sample size datasets (such as microarray data). A fundamental statistical issue is: Which clusters are "really there," as opposed to being artifacts of the natural sampling variation? SigClust is proposed as a simple and natural approach to this fundamental statistical problem, and its properties are studied.
|
