|
Activity Number:
|
119
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 7, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Nonparametric Statistics
|
| Abstract - #305763 |
|
Title:
|
Nonparametric Mixture Model
|
|
Author(s):
|
Mian Huang*+
|
|
Companies:
|
The Pennsylvania State University
|
|
Address:
|
303 Farmstead Lane, State College, PA, 16803,
|
|
Keywords:
|
local likelihood ; mixture of regression ; EM algorithm
|
|
Abstract:
|
Methods of mixture model deal with data where observations are from several homogeneous subgroups. When the data has two or more dimensions and there are relationships between the responses and predictors, mixture of regression model is proper to use. The model is extended to a general case so that the nonlinear relationship in each component can be estimated. The estimation is based on optimizing the local likelihood function using EM algorithm. Simulation studies show that the model had some advantages.
|
