clustering

It is an unsupervised machine learning technique that attempts to group unlabelled data. The most common approach is to use K-means although there other other methods as well. One of the limitations of K-means is trying to determine the optimum number of clusters. Other techniques like G-means handle this limitation differently.