|
Activity Number:
|
294
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Nonparametric Statistics
|
| Abstract - #309035 |
|
Title:
|
Testing Equality of Mean Functions for Continuous Time Stochastic Processes
|
|
Author(s):
|
Yolanda Munoz*+
|
|
Companies:
|
University of Texas at Houston
|
|
Address:
|
1200 Herman Pressler, Houston, TX, 77030,
|
|
Keywords:
|
u-statistic ; non-parametric methods ; jackknife ; funtional data analysis ; ANOVA
|
|
Abstract:
|
A test statistic is developed to determine whether two or more groups of functions, derived from continuous time stochastic processes, belong to the same population. An approximate chi-square test is proposed for comparing n group means. The test statistic considered in this paper uses a quadrature approximation to the L_2 norm. The proposed test statistic is shown to be a U-statistic and its asymptotic properties are formulated. A small Monte Carlo study is conducted and we illustrate its application with a real data set.
|
