K-means algorithm is one of the most commonly used partition based clustering algorithms. K-means has recently been predictable as one of the best algorithms for clustering large data. The formation of empty clusters is one of the most important issues in K-means algorithm. This problem is considered insignificant when the data set is small and can be solved by executing the algorithm for a number of iterations. In some cases, the K-means is used as an essential part in some scientific applications like medical database; the empty cluster problem may affect the behavior of the system along with the performance of the algorithm. In this research article we propose a framework for enhancing the efficiency of K-means algorithm to avoid the formation of empty clusters using data structure. Experimental results show that the enhanced method can effectively improve the speed of clustering, accuracy and avoiding the formation of empty clusters.

Keywords: clustering; k-means Algorithm; Enhanced k-means Algorithm