|
Activity Number:
|
149
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Nonparametric Statistics
|
| Abstract - #309090 |
|
Title:
|
Smoothed Mann-Whitney Method for Two-Sample Location Problem
|
|
Author(s):
|
Feridun Tasdan*+
|
|
Companies:
|
Western Illinois University
|
|
Address:
|
Department of Mathematics, Macomb, IL, 61455,
|
|
Keywords:
|
Shift Parameter ; Robust Estimation ; Two Sample Problem ; Location Parameter
|
|
Abstract:
|
This study is mainly concerned about estimating a shift parameter in the two-sample location problem. Anderson and Hettmansperger (1996) proposed a new method called Generalized Wilcoxon which uses a nondecreasing, and odd function $\psi$ for the pairwise differences of residuals. The proposed Smoothed Mann-Whitney method uses convolution technique for smoothing the empirical distribution functions of the each sample. Then it replaces the empirical cumulative distribution functions with these new smoothed distribution functions. The unknown shift parameter is estimated by solving gradient function with respect to an arbitrary shift variable. Asymptotic properties of the new estimator are established under the conditions that are similar to those used in Generalized Wilcoxon.
|
