Properties of the estimated best linear unbiased predictor (EBLUP) under infill asymptotics in the spatial linear model /
This dissertation investigates the effects of infill-
| Main Author: | |
|---|---|
| Format: | Thesis Book |
| Language: | English |
| Published: |
[Place of publication not identified] :
[publisher not identified] ;
1995.
|
| Subjects: | |
| Online Access: | http://proxy.library.tamu.edu/login?url=http://proquest.umi.com/pqdweb?did=742745341&sid=1&Fmt=2&clientId=2945&RQT=309&VName=PQD |
| Summary: | This dissertation investigates the effects of infill- asymptotics on properties of the spatial best linear unbiased predictor (BLUP) and estimated BLUP (EBLUP). By exploiting relationships between the mean squared prediction error (MSPE) of the BLUP and the EBLUP, it is possible to infer the limiting behavior of the EBLUP's MSPE. These properties were studied under several one- and two-dimensional spatial configurations. General results on the limiting value of the EBLUP's MSPE were derived and a sufficient condition for the difference between a linear and non-linear covariance to be asymptotically zero were derived. Computer simulations and an extensive analysis of a gridded oceanographic/meteorological data set illustrate the results. |
|---|---|
| Item Description: | Vita. "Major Subject: Statistics". |
| Physical Description: | xii, 99 leaves : illustrations ; 28 cm. Issued also on microfiche from University Microfilms Inc. |
| Bibliography: | Includes bibliographical references. |