|
Activity Number:
|
325
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
IMS
|
| Abstract - #308824 |
|
Title:
|
Clustering Time Course Gene Expression Data Using Nonparametric Hypothesis Testing
|
|
Author(s):
|
Haiyan Wang*+ and James Neill and Forrest Miller
|
|
Companies:
|
Kansas State University and Kansas State University and Kansas State University
|
|
Address:
|
108 E Dickens Hall, Manhattan, KS, 66503,
|
|
Keywords:
|
Clustering ; Hypothesis testing ; functional data ; Multiple testing ; High dimensional data ; asymptotics
|
|
Abstract:
|
This paper presents an agglomerative clustering method for effectively detecting patterns and clusters in high dimensional functional data. The algorithm is based on two new nonparametric hypothesis tests applicable to curve screening and comparing groups of curves with heteroscedastic nonstationary observations over time. Flat curves are first screened and then two sets of curves are clustered into one group if they are not significantly different. The number of clusters are determined automatically by given significance levels. Our procedure takes into account both shape and magnitude of curves, is invariant under monotone transformation of data, and does not suffer from information loss due to the use of smoothing based procedures. Simulations show that the algorithm has low clustering error rates.
|
