A Grid and Density Based Fast Spatial Clustering Algorithm
Ming Huang, Fuling Bian · 2009
Density-based spatial clustering algorithm DBSCAN has a relatively low efficiency since it carries out a large number of useless distance computing; Grid-based spatial clustering algorithm is more efficient, but the clustering result has a low accuracy. Considering the advantage and disadvantages of the two algorithms, this paper proposes a grid and density based fast clustering algorithm GNDBSCAN. This algorithm performs density-based clustering on datasets space, which has been divided by grids. It improves the efficiency of clustering and at the same time, maintains high accuracy for clustering results.