A new clustering algorithm that claims to be much faster than existing clustering algorithms while still producing very good results. Unlike K-Means it does not require the number of clusters to be given a priori. (tags: machinelearning programming code algorithm)