|
Activity Number:
|
120
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Nonparametric Statistics
|
| Abstract - #305634 |
|
Title:
|
A Multivariate Median Based on the Minimum Spanning Tree
|
|
Author(s):
|
Amber E. Weyand*+ and Adam Petrie
|
|
Companies:
|
The University of Tennessee and The University of Tennessee
|
|
Address:
|
SMC339, Knoxville, TN, 37996,
|
|
Keywords:
|
multivariate median ; robust estimator of location ; depth ; spanning tree
|
|
Abstract:
|
There is no unique way to determine a multivariate median. Several methods have been proposed in past years. Some of these are depth-based estimators, while others are based on graph theory. An overview of such procedures will be provided. Also, we will also present a new technique which utilizes the minimum spanning tree. We will compare the performance of this new procedure with depth-based methods on both artificial and real data sets. Specific details of the advantages and disadvantages of the methods will be discussed.
|
