ABSTRACT
This chapter introduces the model-based clustering is related to standard heuristic clustering methods and an overview of different ways to specify the cluster model. It provides the clustering problem, strategies for selecting a suitable mixture model together with the appropriate clustering base. The chapter discusses how to obtain an identified model, derive a partition of the data, characterize the cluster distributions, gain insights through suitable visualizations and validate the clustering. It explores the model-based clustering has been used in a range of different applications from which also methodological advances have emerged. Mixture models extend the toolbox of clustering methods available to the data analyst. Heuristic clustering methods are based on the definition of similarities or dissimilarities between observations and groups of observations. In hierarchical clustering some linkage methods, that is, distance definitions between groups of observations, also lead to solutions with high compactness, such as complete linkage.
