seminario

[Pietro Muliere <pietro.muliere@uni-bocconi.it>]

Istituto Metodi Quantitativi - Università L. Bocconi
Viale Isonzo, 25 - 20135 Milano
Tel. 02-58365629  - Fax 02-58365630




SEMINARIO



Bayesian Approaches for Multivariate Spatial Process Modeling 
using  Coregionalization

Alan E. Gelfand,
Duke University


Giovedì, 21 novembre ore 16.30
Aula IMQ  stanza n.137

___________________________________________________________________________
Abstract . Models for the analysis of spatial data are receiving increased 
attention these days.  In many applications it will be preferable to work 
with multivariate process specifications in providing such models.  A 
critical specification in developing these models is the cross-covariance 
function.  An attractive constructive approach for creating rich, 
computationally manageable classes of such functions is through the linear 
model of coregionalisation (LMC).The contributions of this paper include: 
fully Bayesian development of  the LMC including the posterior distribution 
of component ranges; clarification of the connection between joint and 
conditional approaches to the fitting of such models including prior 
specifications; extension of  the LMC, including spatially varying LMC's 
and connections with other constructive approaches. Various computational 
issues arise and we discuss them.  Also, we provide several examples 
including the analysis of a particular day average of CO, NO, and NO2 for a 
set of monitoring stations in California and the analysis of a set of 
commercial property transactions in the city of Chicago. This is joint work 
with Alexandra M. Schmidt and C.F. Sirmans.