A Selective Fuzzy Clustering Ensemble Algorithm

To improve the performance of clustering ensemble method, a selective fuzzy clustering ensemble algorithm is proposed. It mainly includes selection of clustering ensemble members and combination of clustering results. In the process of member selection, measure method is defined to select the better clustering members. Then some selected clustering members are viewed as hyper-graph in order to select the more influential hyper-edges (or features) and to weight the selected features. For processing hyper-edges with fuzzy membership, CSPA and MCLA consensus function are generalized. In the experiments, some UCI data sets are chosen to test the presented algorithm’s performance. From the experimental results, it can be seen that the proposed ensemble method can get better clustering ensemble result.