ABSTRACT

The existing clustering algorithm has a sequential execution of the data. The speed of the execution is very less and more time is taken for the execution of a single data. A new algorithm Parallel Implementation of Genetic Algorithm using KMeans Clustering (PIGAKM) is proposed to overcome the existing algorithm. PIGAKM is inspired by using KM clustering over GA. This process indicates that, while using KM algorithm, it covers the local minima and it initialization is normally done randomly, by KM and GA. It always converge the global optimum eventually by PIGAKM. To speed up GA process, the evalution is done parallely not individually. To show the performance and efficiency of this algorithms, the comparative study of this algorithm has been done.

Keywords: - Clustering, Genetic algorithm, K-means, Mutation , Parallel.