A survey analysis on machine learning techniques big data analytics

Abstract

Big data is tremendously developing in all the fields of science and engineering. The potential of enormous information is much essential to introduce innovative thinking and robust methods for addressing different challenges. Analyzing big data enables the optimization process thereby empowering decision making and insight discovery.  The study deals with analyzing the modest trends in the processing of big data with the aid of machine learning. The Big data characteristics process the data visualization as a critical task since the abstract view will not provide accurate geo representation of the data. In particular, the algorithms employed in the big data tools should identify trends, patterns and correlations. Four main challenges of the big data which are veracity, variety velocity and volume have also been considered. Our review investigated the big data analytical methods with respect to the nature of the data with outgrowing learning technologies in recent years like transfer learning, parallel and distributed learning, active learning, representation learning and kernel-based learning. The big data applications in health care, e-commerce, education system etc. were deeply discussed in the study. The review also focuses on the discussions and analytical matrix followed by open issues regarding the existing challenges and reliable solutions for big data processing with machine learning that can benefit learners, educators and administrators.