Cluster Analysis of Spatial Data Using Peano Count Tree

摘要

Cluster Analysis of spatial data is a very important field due to the very large quantities of spatial data collected in various application areas. In most cases the spatial data sizes are too large to be mined in a reasonable amount of time using existing methods. In this paper, we restructure the data using a data-mining ready data structure, Peano Count Tree (P-tree) and then based on this new data formats, we present a new scalable clustering algorithm that can also handle noise and outliers effectively. We achieve this by combining the partitioning and density-based clustering methods. Our analysis shows that with the assistance of P-trees, very large data sets can be handled efficiently and easily.