|
Activity Number:
|
141
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 4, 2008 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Nonparametric Statistics
|
| Abstract - #301632 |
|
Title:
|
A Nonparametric Test of Independence Between Response and Covariate Adjusted for Treatment Effect
|
|
Author(s):
|
Siti Tolos*+ and Haiyan Wang
|
|
Companies:
|
Kansas State University and Kansas State University
|
|
Address:
|
Dickens 101, Manhattan, KS, 66506,
|
|
Keywords:
|
Nonparametric covariate effects ; Fully nonparametric model ; Nearest neighborhood ; Fixed window size ; Measure of dependency
|
|
Abstract:
|
Effective methods for detecting general dependency among variables are in high demand, especially in high dimensional post genome data due to the complex nature of the data. Existing methods mainly focused on testing for mutual independence. In this paper, we will discuss a nonparametric test of independence between response and covariates after adjusting for nonparametric treatment effects. The test statistic is constructed using moment methods with a fixed number of nearest neighbors as pseudo replicates. The asymptotic distribution of the proposed test statistic will be presented, followed by simulation studies. Potential applications in obtaining initial connected graphs for constructing gene regulatory networks will be discussed.
|
