Hybrid Fuzzy C-Means Clustering Algorithm Oriented to Big Data Realms

A hybrid variant of the Fuzzy C-Means and K-Means algorithms is proposed to solve large datasets such as those presented in Big Data. The algorithm sensitive initial values membership matrix. Therefore, a special configuration matrix can accelerate convergence algorithm. In this sense, new approach proposed, which we call Hybrid OK-Means (HOFCM), it optimizes parameter. This consists three steps: (a) generate set n solutions an x dataset, applying algorithm; (b) select best solution basis for generating optimized matrix; (c) resolve dataset with C-Means. experimental results four real one synthetic show that HOFCM reduces time by up 93.94% compared average standard It highlighted quality was reduced 2.51% worst case.