|
Activity Number:
|
324
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 2, 2016 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Nonparametric Statistics
|
|
Abstract #318674
|
|
|
Title:
|
Depth Functions and Classification Using Beta-Skeleton Graphs
|
|
Author(s):
|
Yu Song*
|
|
Companies:
|
The George Washington University
|
|
Keywords:
|
beta-skeleton ;
U-statistic ;
high dimension ;
distribution function ;
interpoint distance
|
|
Abstract:
|
We define the beta-skeletons depths based on the probability that a point is contained within the beta -skeleton influence region of two i.i.d. random vectors. We show that the -skeleton depths are a family of statistical depth functions. We define and examine the sample beta -skeleton depth functions and show that they have desirable asymptotic properties. We also explore the beta -skeleton depth to test for the equality of the high dimensional distribution functions and classi cation. A Monte Carlo study compares the beta -skeleton test with some existing statistics.
|
Authors who are presenting talks have a * after their name.
