A Fuzzy Clustering Ensemble Based on Dual Boosting*

Clustering ensemble is fit for any shape and distribution datset . Boosting methodogy provides superior results for classification problems. In the paper, A dual boosting is proposed for ensemble of fuzzy clustering . At boosting iteration , a new training set is created based on the original datasets´ weights which is associated with the previous clustering . According the dual boosting method, the new training set not only include the datas which is hard to clustering ,but also includes the dta which is easy to cluster . The final clustering solution is propuced by re-clustering based on the co-association matrix. Experiments on both artifical and real word data sets indicate that the dual boosting clustering ensemble provides solutions of improved quality.