Performance evaluation of cluster algorithms for Big Data analysis on cloud
Chu‐Hsing Lin, Jung‐Chun Liu, Tsung-Chi Peng · 2017
In this study, based on clustering algorithms, we perform data mining on land price data in Taichung City during past ten years. For big data analysis, we combine Hadoop HDFS and MapReduce with R language and visualize results on Google Maps. We also study performances of K-means and Fuzzy C-means clustering algorithms, executed in the Hadoop cloud and a stand-alone PC. The experimental results show that with a cloud of 9 compute nodes, about 3.5 times of acceleration are attainable; hence Hadoop cloud with R can be applied to solving insufficient memory issues in big data applications.