Activity Number:
|
492
- Application of Nonparametric Methods
|
Type:
|
Contributed
|
Date/Time:
|
Thursday, August 6, 2020 : 10:00 AM to 2:00 PM
|
Sponsor:
|
Section on Nonparametric Statistics
|
Abstract #312397
|
|
Title:
|
Spatio-Temporal Single Index Models for Correlated Data
|
Author(s):
|
Hamdy Mahmoud* and Inyoung Kim
|
Companies:
|
Virginia Tech and Virginia Tech
|
Keywords:
|
MCEM algorithm;
mixed model;
single index model;
spatio-temporal data
|
Abstract:
|
Modeling spatially-temporally correlated data using parametric models is common, however semiparametric models are very limited in this area. This article introduces a semiparametric spatial-temporal effects model, in which spatial effects are integrated into the single index function and temporal effects are additive to the single index function. We refer to this model as ``semiparametric non additive spatio-temporal single index model" (NST-SIM). For estimation, Monte Carlo Expectation Maximization based algorithm is used. NST-SIM has many advantages demonstrated by simulation studies and real data application. It has smaller mean square error and higher accurate prediction compared to non-integrated spatial temporal single index model. It is applied to South Korean mortality data of six major cities and interesting results are found.
|
Authors who are presenting talks have a * after their name.