Clustering analysis

Clustering analysis is a type of unsupervised learning which aims to find the most natural way of grouping a dataset. This is achieved by organizing a set of observations based on a similarity criterion, such that observations in the same group are more alike than observations in different groups. In this chapter we use K-means, the most popular clustering algorithm, to illustrate the main concepts, advantages, and limitations of clustering analysis. We then present alternative clustering algorithms including Gaussian mixture model and density-based spatial clustering of applications with noise. Finally, we illustrate some applications of K-means to the investigation of brain disorders and conclude with a series of recommendations.