A fuzzy approach to intensive data mining

Abstract

For many spatial processes, there is a natural need to find out the point of origin on the basis of the available scatter of observations; think for instance of finding out the home-base of a criminal given the actual distribution of crime scenes, or the outbreak source of an epidemics. We introduce a new methodology based on the notion of Topological Weighted Centroid (TWC) that allows one to draw powerful inferences also in relatively intractable cases with few observations or a poorly understood underlying data generating process. In this paper we consider reconstruction of global political and economic relationships on the basis of a small-dimensional qualitative dataset.