Cluster identification

Cluster analysis is an important and frequently-used data mining tool with a wide variety of research applications. There are numerous methods by which clusters in a dataset can be identified, with the majority of tools used for identifying statistical clusters in feature space. There also exists a specific subset of clustering approaches appropriate for the identification of spatial and spatiotemporal clusters with geographic data. The theory of these methods is presented here, as are common pitfalls and parameterization methods. This chapter concludes with a detailed discussion on how several different common spatial and spatiotemporal cluster analyses can be applied to different datasets.