Application of spatial data mining methods for the definition of a predictive urban growth model

Abstract

This research aims at the analysis of urban region development and more specifically, the determination of an urban model, concerning the diachronic land use changes. Under this perspective, Data Mining methods and techniques are applied, for the spatial analysis and clustering of corresponding data at the level of municipality (fuzzy clustering) and for the detection of future trends (neural network model). The proposed methodological approach, a model is being built for a twenty years forecast, the theoretical background of which is that the degree of spatial adjacency between the individual municipalities, is a decisive factor for their future evolution.