A Hybrid Collaborative Clustering Using Self-Organizing Map

Abstract

In this study, we introduce a novel hybrid collaboration clustering architecture, in which several subsets of patterns can be processed together with an objective of finding a common structure. The structure revealed at the global level is determined by exchanging prototypes of the subsets of data and by moving prototypes of the corresponding clusters toward each other. Thereby, it comprises a judicious integration of the principles of vertical and horizontal collaboration using the Self-Organizing Map (SOM). A detailed clustering algorithm is developed by integrating the advantages of both collaboration clustering. The effectiveness of the algorithm, along with a comparison with other algorithms, has been demonstrated on a set of real life data sets. The power of collaboration between every pair of datasets is estimated by a parameter, we call coefficient of collaboration, to be determined iteratively during the collaboration phase using a steepest descent method based optimization, for the algorithm. Promising results discovered the deep impact observed at the individual clusters, permitting us to conclude that the global effect of the collaboration has been ameliorated. The proposed method has been validated on several datasets and experimental results have presented very promising performance.