Differently from hierarchical clustering procedures, non-hierarchical clustering methods need the user to specify in advance the number of clusters; therefore, in this case, a single partition is obtained. The two most famous non-hierarchical clustering algorithms are the k-Means and the k-Medoids one. They differ in the definition of the cluster prototypes. In particular, the k-Means prototypes, called centroids, are defined to be the average values of units assigned to the clusters, while the k-Medoids prototypes, called medoids, identify the most representative observed units for each cluster. In this chapter, non-hierarchical clustering methods will be briefly introduced from a theoretical point of view and their implementation will be presented in detail by means of some real-life case studies.
CITATION STYLE
Giordani, P., Ferraro, M. B., & Martella, F. (2020). Non-Hierarchical Clustering (pp. 75–109). https://doi.org/10.1007/978-981-13-0553-5_3
Mendeley helps you to discover research relevant for your work.