Abstract:
|
Spatial data analysis has received increasing attention in several applications, and spatial data have been progressively observed and recorded in various forms. Symbolic data is defined as a hypercube in p-dimensional space and embraces a variety of data forms, such as histograms, lists, and intervals. In this paper, we specifically focus on spatial interval-valued data, which is observed in the form of intervals over space. We propose a statistical framework for spatial interval-valued data analysis in order to study spatial structure and to conduct spatial prediction at unobserved locations. A simulation study is performed to compare the proposed methods and a real data set is used to illustrate the methods. Several novel evaluation measures are proposed to evaluate the prediction accuracy.
|