New improved technique for initial cluster centers of K means clustering using Genetic Algorithm

Cluster Analysis is one of the most important data mining techniques which help the researchers to analyze the data and categorize the attributes of data into various groups. K-Means is one the frequent partitioning algorithm used in clustering. The enhancement of K-means clustering can be done by choosing appropriate initial cluster centers to converge quickly to the local optimum. In the proposed work, I intend to choose the initial cluster centers using Genetic Algorithm instead of choosing them randomly which would lead us to improved solutions and decreased complexity of the conventional k-means algorithm. The paper suggests that initialization of the cluster centers cannot be separated from effectiveness and the concept of success and failure. The randomness of the cluster centers need to be managed and controlled so as to put a limit on the number of iterations to be carried out in the conventional algorithm with decreased complexity and increased accuracy.