Semi-supervised Affinity Propagation clustering

embed Silhouette index into iterations of Affinity propagation clustering to supervise its running

Affinity propagation clustering (AP) is a clustering algorithm proposed in "Brendan J. Frey and Delbert

Dueck. Clustering by Passing Messages Between Data Points. Science 315, 972 (2007)". It has some advantages: speed, general applicability, and suitable for large number of clusters.

Semi-supervised AP improves AP by: embedding Silhouette indices into the programs of AP to supervise the running of AP, so that the AP will give its optimal clustering solution.

The programs of semi-supervised AP are suitable for the person who has interests in studying or improving AP algorithm, and then the semi-supervised AP may be an example for reference.