Activity Number:
|
474
- Nonparametric Density and Variance Estimation
|
Type:
|
Contributed
|
Date/Time:
|
Wednesday, August 2, 2017 : 8:30 AM to 10:20 AM
|
Sponsor:
|
Section on Nonparametric Statistics
|
Abstract #323942
|
View Presentation
|
Title:
|
Orthogonal Series Density Estimation for Complex Surveys
|
Author(s):
|
Shangyuan Ye* and Ye Liang and Ahmad A. Ibrahim
|
Companies:
|
Oklahoma State University and Oklahoma State University and Oklahoma State University
|
Keywords:
|
Nonparametric ;
asymptotic ;
survey sampling ;
orthogonal basis ;
mean integrated squared error
|
Abstract:
|
Density estimation is an important topic in survey sampling since it summarizes all relevant information about the population of interest. However, the assumption of independent and identically distributed samples are violated when data are sampled from a finite population using a complex sampling design. In this paper, we propose an orthogonal series density estimation (OSDE) for complex surveys. The proposed estimator is design-unbiased and asymptotically design-consistent. The asymptotic normality is studied under both design and combined spaces. We also show the lower bound of minimax mean integrated squared error (MISE) for the proposed estimator. Data driven estimators are proposed and studied via both simulation and real data analysis.
|
Authors who are presenting talks have a * after their name.