Large Scale Machine Learning with Python

Clustering – K-means

K-means is an unsupervised algorithm that creates K disjoint clusters of points with equal variance, minimizing the distortion (also named inertia).

Given only one parameter K, representing the number of clusters to be created, the K-means algorithm creates K sets of points S₁, S₂, …, S_K, each of them represented by its centroid: C₁, C₂, …, C_K. The generic centroid, C_i, is simply the mean of the samples of the points associated to the cluster Si in order to minimize the intra-cluster distance. The outputs of the system are as follows:

1. The composition of the clusters S₁, S₂, …, S_K, that is, the set of points composing the training set that are associated to the cluster number 1, 2, …, K.
2. The centroids of each cluster, C₁, C₂