An enhanced K-means genetic algorithms for optimal clustering

K-means algorithm is sensitive to the initial cluster centers and clustering results diverge with different initial input which in turn falls into local optimum. Genetic Algorithms are randomized searching technique which provides a better optimal solution for fitness function of an optimization problem. This paper proposes an enhanced K-means Genetic Algorithm for optimal clustering of data (EKMG). The aim is to maximize the compactness the clusters with large separation between at least two clusters. The superiority of EKMG is compared with grouping genetic algorithm (GGA) by using real-life dataset. The experiment shows that EKMG reaches better optimal solution with high accuracy.