EFFICIENT BIG DATA CLUSTERING USING A HYBRID K-MEANS WITH CLUSTER MERGING STRATEGY

Abstract

 The use of big data is currently exploding across the board in academia and industry. As essential as the potential of this huge data is, new methods of thinking and training are needed to overcome many of the problems. This vast prospective of the data is undeniable, thus different methods of thinking and new learning methodologies are needed for solving these many issues. Clustering techniques are the most important issues to handle big data efficiently. Many clustering techniques are used to simplify the way we manipulate Big data. Despite the fact that k-means is one of the most often used clustering approaches, many researchers introduced modifications to increase its efficiency. Several frameworks are designed to facilitate the manipulation of big data, Map Reduce is one of the most useful frameworks used. In this paper, we present a modified k-means algorithm based on mergeing nearest clusters that increased its efficiency according to our simulation on benchmark data.