Evolution of Data Analytics Techniques: From Data Mining to Big Data Mining

  • Gourav Bathla, Himanshu Aggarwal, Rinkle Rani

Abstract

Data mining is one of the most significant techniques for researchers, business organizations and decision analyst. It has helped data analyst in improving business model, healthcare, weather forecasting, e-mail spam filtering, social networks and many other domains. It has evolved in recent years at a very high pace. The motive of this paper is to highlight different phases of data mining evolution in frequently changing requirements. Specifically, ten algorithms which are identified as most prevalent by researchers are studied and observed in this paper. K-means, SVM and Naïve Bayes are compared based on several characteristics. Moreover, Big data mining, machine learning and deep learning techniques which are advanced form of data analytics are observed with their specific applications. Graphics Processors (GPU) which is essentialfordeeplearningtechniquesareanalyzedindetailinthispaper.Several research works have discussed the advantages and limitations of these data analytics techniques. However, to the best of our knowledge very few works have provided comprehensive analysis of these techniques. Readers can identify limitations and advantages of these techniques which are evolved over the time and select the best technique which is most suitable for their research or applications.

 

Keywords: Data mining, Classification, Clustering, Big data, Social recommendation, Machine learning, Deep learning, GPU

Published
2019-12-31
How to Cite
Rinkle Rani, G. B. H. A. (2019). Evolution of Data Analytics Techniques: From Data Mining to Big Data Mining. International Journal of Advanced Science and Technology, 28(19), 779 - 795. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/2664
Section
Articles