A geostatistical model for spatial point aggregated data
Francesco Finazzi (University of Bergamo)
Friday 12th October, 2012 16:15-17:15 Maths 203
This seminar addresses the problem of estimating a spatially continuous random process from spatial point aggregated data which are supposed to carry information on the process itself. In particular, spatial point aggregated data are georeferenced to precise points in space but they refer to sets of statistical units.
Since spatial point aggregated data exhibit interaction, neither the classic geostatistical models nor the classic spatial point process models can be adopted to estimate the realization of the underlying random process. Therefore, an approach based on a novel geostatistical model is developed.
As a case study, the estimation of the spatial market potential of a newspaper over the area of a city is considered. The sales data of the newsstands in the city are used to estimate the spatial market potential which is eventually analyzed in order to identify the areas with the highest market potential.