It clusters data based on distance to a cluster center (as in K-means), but also detects non-spherical clusters (DBSCAN). This new method is able to efficiently identify the correct number of ...