|
Activity Number:
|
244
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 1, 2016 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics and the Environment
|
|
Abstract #318633
|
|
|
Title:
|
Spatial Data Fusion for Large Non-Gaussian Remote Sensing Data Sets
|
|
Author(s):
|
Hongxiang Shi* and Emily Lei Kang
|
|
Companies:
|
University of Cincinnati and University of Cincinnati
|
|
Keywords:
|
EM algorithm ;
Empirical Bayes ;
Hierarchical Modeling ;
Multivariate geostatistics ;
Spatial random effects model
|
|
Abstract:
|
Remote sensing data are playing a vital role in understanding the pattern of the Earth's geophysical processes. In this paper, we proposed a spatial data fusion method that is able to take advantage of two (or potentially more) large remote sensing datasets with the exponential family of distributions. We take an empirical Hierarchical modeling (EHM) approach where any unknown parameters are estimated by Maximum Likelihood estimation via an efficient EM algorithm. Then through a MCMC algorithm the prediction is obtained by generating samples from the empirical predictive distribution where the unknown parameters are substituted by the estimates from the EM algorithm. Finally, the performance of our proposed method are investigated through simulation studies and real datasets. It shows that our data fusion method has capabilities to borrow strength across two complementary datasets and thus improving predictions reciprocally.
|
Authors who are presenting talks have a * after their name.
