Research on Spatially Weighted Fuzzy Dynamic Clustering Algorithm and Spatial Data Mining Visualization

Abstract

Generally the spatial fuzzy clustering algorithm has not differentiated between the various attributes imbalance and the multi-dimensional data description complexity and that isn’t favor to the complex spatial data relation correct description. In response to this problem, this paper uses a spatially weighted fuzzy dynamic clustering algorithm and realizes the dynamic process visualization of the algorithm. The research combines the weight of AHP with spatial fuzzy clustering and then uses statistical probability F-distribution to determine the optimal threshold in order to enhance the intelligence of spatial fuzzy clustering algorithm. The use of Star coordinate can achieve the spatial fuzzy clustering algorithm visualization. Tests show that compared with based on fuzzy equivalent relation transitive closure method, the clustering is higher than the not weighted fuzzy clustering algorithm.