AbstractClustering is one of the widely used data mining techniques. In machine learning, it is an unsupervised learning method that needs training data to be used. The problem with clustering is that it is computationally expensive and takes more time for grouping high-dimensional data. Therefore it is necessary to reduce number of features in the high-dimensional data in order to make the search space reduced. Many techniques came into existance for feature selection. However, they are not efficient for high-dimensional data. In this paper, we proposed a methodology to build an approach that focuses on choosing representative features that can be used for making clustering decisions instead of using all features. Thus the search space is reduced, thereby improving speed and computational efficiency. We proposed an algorithm to achieve this. We built a prototype application to show proof of the concept. The empirical results revealed that the proposed algorithm is effective as it exploits representative features.Keywords: Data mining, clustering, feature selection, high-dimensional dataCite this ArticleVamsi Krishna G, Seepana Priyanka. Representative Feature Selection for Efficient Clustering. Journal of Network Security. 2017; 5(2): 19–27p.
