The proposed Center Initialization Based on Imperialist Competitive Algorithm (CIB-ICA)

In this paper we will introduce an initial cluster centers method for k-means algorithm, which can achieve a significant impact on the convergence and will not fall local optimal solution trap .The proposed Center Initialization Based on Imperialist Competitive Algorithm (CIB-ICA) uses imperialist competitive algorithm and minimum spanning tree(MST) features to reduce clustering error percentage of K-means algorithm. The proposed method has been evaluated on some famous datasets and experimental results show that it is an efficient cluster center initialization method.