Cluster-Based Ensemble Using Distributed Clustering Approach for Large Categorical Data

Abstract

The modern data analytics approaches suffer through the various challenges when dealing with large datasets, and thus, there is qualitative and quantitative degradation in the decision-making process. There is a need of an efficient analytical system that can handle on large and diverse datasets with increased accuracy in data analysis. Here, a prototype model of a cluster-based ensemble system is proposed to run on commodity machines using hierarchical approach. The commodity machines create the cluster-based ensemble in a distributed environment using a popular data mining algorithm. The experimentation proves that hierarchical approach helps to decrease time for data mining process, in turn boosts the speed of decision-making and the cluster-based ensemble improved the accuracy.