An incremental grid clustering algorithm based on density-dimension-tree

This paper proposes an approach to improve the existing grid-based clustering algorithms with a further grid partition strategy and an incremental clustering function. This new algorithm IGDDT is based on density-dimension tree, which has the ability to reuse the previous clustering results, and obtain the better clusters by further dividing the grid cell in the clustering process. The experimental results on both artificial and real datasets demonstrate that IGDDT is able to discover arbitrary shape of clusters, better performance than the previous clustering algorithms on both clustering accuracy and clustering efficiency.