Fast Affinity Propagation Clustering based on Machine Learning
Shailendra Kumar Shrivastava, J. L. Rana, Ravindra Jain, Samrat Ashok · 2013
Affinity propagation (AP) was recently introduced as an unsupervised learning algorithm for exemplar based clustering. In this paper a novel Fast Affinity Propagation clustering Approach based on Machine Learning (FAPML) has been proposed. FAPML tries to put data points into clusters based on the history of the data points belonging to clusters in early stages. In FAPML we introduce affinity learning constant and dispersion constant which supervise the clustering process. FAPML also enforces the exemplar consistency and one of ‘N’ constraints. Experiments conducted on many data sets such as Olivetti faces, Mushroom, Documents summarization, Thyroid, Yeast, Wine quality Red, Balance etc. show that FAPML is up to 54 % faster than the original AP with better Net Similarity.