An Efficient Approach towards K-Means Clustering Algorithm

K-Means clustering algorithms are used in various practical applications countless times. Original K-Means algorithm select initial centroids randomly it generates unstable cluster as the value of object in cluster depend on the selection of initial cluster means which is done by random selection of objects. The number of times different selection of initial centroids will give number of different clusters with different accuracy. The algorithm used in base paper’s clustering method eliminates the deficiency of K-Means. It first computes the initial centroids k according to the requirements of users and then gives better, effective and good cluster. To improve accuracy it generates stable clusters. It also reduces the mean square error and improves the quality of clustering, but this algorithm has large execution time, which makes it expensive. As this algorithm requires storing lot of calculations it requires lot of space The proposed algorithm that is the new efficient approach towards K-Means algorithm combines the method of both the algorithm it systematically chooses the initial centroids for the procedure in such a way that it reduces mean square error without a high increase execution time. This algorithm requires less space than the base paper’s algorithm.