Single link We now consider the single link or MIn version of hierarchical clustering. In this case, the proximity of two clusters is defined as the minimum of the distance between any two points in the two different clusters This technique is good at handling non-elliptical shapes However, it is sensitive to noise and outliers 8/20/2016 PATTERN RECOGNITIONSingle link We now consider the single link or MIN version of hierarchical clustering. In this case, the proximity of two clusters is defined as the minimum of the distance between any two points in the two different clusters. This technique is good at handling non-elliptical shapes. However, it is sensitive to noise and outliers. 8/20/2016 PATTERN RECOGNITION 49