Point and Functional Entropy-Based Spatial Sampling Design
by
Jose M. Angulo
Department of Statistics and O.R., University of Granada, Spain
Coauthors: Maria D. Ruiz-Medina (Department of Statistics and O.R., University of Granada, Spain), Francisco J. Alonso (Department of Statistics and O.R., University of Granada, Spain), Maria C. Bueso (Department of Statistics and O.R., University of Granada, Spain)

Entropy-based criteria provide a meaningful approach to obtaining optimal designs of point observation strategies for the prediction of spatial processes. The practical procedures derived are particularly suitable to implementation in the case of Gaussian processes. Extensions to space-time processes and to multivariate situations offer interesting research directions in this field. An important generalization consists of functional optimal design, that is, optimal selection of test functions to extract average information from a given random field. The theory of generalized random fields provides a convenient framework to develop such a generalized approach. A relevant application consists of optimal selection of truncation schemes in orthogonal expansions, leading to optimal finite-discretization in the form of specific regularizations of a random field. The object of this talk is to present a review and discussion of such problems, which are of particular interest in the field of Environmental Statistics research and applications.

Date received: November 15, 2000

Copyright © 2000 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cafx-07.