Abstract: The term big data has been used in recent years to describe the enormous volume of data collected by various means and at a very rapid pace in all areas of human activity. Big data differs from traditional data sets and requires special management due to its volume, structure and type. For this reason, specialized technologies for their processing are being developed. The use of machine learning methods to take advantage of the big data collected in the health sector is imperative as it can significantly contribute to making predictions, preventing illnesses, improving patient healthcare, reducing costs and optimizing the use of available resources. This paper deals with Big Data, briefly presenting its characteristics, the sources of its creation, as well as the basic technologies used for its processing and utilization in the health sector. In the second part of this work, medical data on diabetes were used, different software tools were integrated to get information using SQL queries and machine learning methods were applied using the Weka open-source software to evaluate and compare the results for diabetes prediction. The paper concludes with useful concepts arising from the analysis, as well as challenges and suggestions for future research.

Keywords: Big Data, Data analytics, Machine Learning, Healthcare, Spark SQL, Weka, Diabetes prediction

Pages: 11-20

WSEAS Transactions on