MKF-Cuckoo: Hybridization of Cuckoo Search and Multiple Kernel-based Fuzzy C-means Algorithm

Discovering of optimal cluster through the help of optimization procedure is a recent trend in clustering process. Accordingly, several algorithms have been developed in the literature to mine optimal clusters. Most of the optimization- based clustering algorithms presented in the literature are only focused on the same objective given in the well-known clustering process, k-means clustering. Instead of k-means objective, some more effective objective functions are designed by the researchers for clustering. So, hybridization of those effective objectives with optimization algorithms can lead the effective clustering results. With the aim of this, we have presented a hybrid algorithm, called MKF-Cuckoo which is the hybridization of cuckoo search algorithm with the multiple kernel-based fuzzy c means algorithm. Here, MKFCM objective is taken and the same objective is solved through the cuckoo search algorithm which is one of the recent optimization algorithm proved effective in many optimization problems. For proving the effectiveness of the algorithm, the performance of the algorithm is comparatively analyzed with some other algorithm using clustering accuracy, rand coefficient, jaccard coefficient and computational time with iris and wine datasets. From the results, we can prove that the hybrid algorithm obtained 96% accuracy in iris data and 67% accuracy in wine data.