Mixture Model-Based Clustering
Sixty years after the notion of defining a cluster as a component in a mixture model was put forth, the use of mixture models for clustering has grown into an important subfield of classification. The volume of work within this field over the past decade seems equal to all of that which went before. Accordingly, it is interesting to review work to date as well as looking at the present and ahead to the future. First, the definition of a cluster is discussed along with some historical context for model-based clustering. Then, starting with Gaussian mixtures, the evolution of model-based clustering is traced out starting in 1965 and concluding with work that is currently available only in preprint form.