Using Rough Classification to Represent Uncertainty in Spatial Data
Ola Ahlqvist, Johannes Keukelaar, Arim Oukbir · 1998
This paper explores the possible advantages of using rough set based classification in GIS, to represent uncertainty both in terms of spatial location and of attribute value determination. Rough classification is based on rough set theory, where an uncertain set is specified by giving an upper and a lower approximation. We present some measures that can be applied to such a rough classification. To compare a rough classification with a crisp one, we present an extension to the error matrix formalism. An experiment using rough classification is performed, to demonstrate its viability. In the experiment, we attempt to compare two incompatible vegetation classifications covering the same area.