Correlating Apache Spark and Map Reduce with Performance Analysis using K-Means

ABSTRACT Big Data has for some time been the point of interest for Computer Science devotees around the globe, and has increased much more conspicuousness in the late times with the nonstop blast of information coming about because of any semblance of social media and the mission for tech mammoths to access further analysis of their information. This paper talks about two of the correlation of - Hadoop Map Reduce and the as of late presented Apache Spark ? both of which give a preparing model to breaking down enormous information. Albeit both of these choices depend on the idea of Big Data, their performance differs altogether in view of the utilization case under usage. This is the thing that makes these two alternatives deserving of analysis as for their fluctuation and assortment in the dynamic field of Big Data. In this paper we contrast these two frameworks along and giving the performance analysis utilizing a standard machine learning algorithm for clustering (K-Means).The look into paper additionally thinks about the performance of the start and MapReduce with the parameters like speed, throughput and energy consumption. General Terms Big Data, Machine Learning, K Means