Agglomerative hierarchical clustering based on affinity propagation algorithm

Abstract

Affinity propagation (AP) algorithm doesn't fix the number of the clusters and doesn't rely on random sampling. It exhibits fast execution speed with low error rate. However, it is hard to generate optimal clusters. This paper proposes an agglomerative clustering based on AP (agAP) method to overwhelm the limitation. It puts forward k-cluster closeness to merge the clusters yielded by AP. In comparison to AP, agAP method has better performance and is better than or equal to the quality of AP method. And it has an advantage of time complexity compared to adaptive affinity propagation (adAP).