Problems Concerning the Compact Support for Scalar and Vector--valued Random Fields Defined Over Space and Time
GSP BROWN BAG SEMINAR
University of Castilla-La Mancha, Spain Abstract:
In the last year there has been an increasing interest for the analyisis of scalar and vector--valued random fields defined over space and time. Practitioners often deal with massive data sets for which estimation through maximum likelihood techniques becomes often unfeasible. Using compactly supported covariance function can allow to overcome such problem and tapering techiques have been proposed, in the past, in the context of scalar--valued random fields defined over space only (Du, Zhang and Mandrekar, 2009). We believe that the extension of compact support to space--time as well as to vector--valued random fields represents a considerable challenge from the mathematical viewpoint and an important tool for the
statistical community. In this talk we present a personal selection of the results obtained in the last three years of research.