Predicting Diabetes using Distributed Machine Learning based on Apache Spark

Diabetes mellitus is a long-standing disease. It constitutes a severe challenge for public health worldwide. As stated by the International Diabetes Federation, there are presently about 246 million diabetic people around the world, and this number is anticipated to increase to around 380 million by the year 2025. More than this, 3.8 million death cases occur annually due to diabetes complications. The primary objective of this work is developing an applicable system to predict diabetes using distributed machine learning based on big data platforms such as Spark. In this context, this study aims to develop models using distributed machine learning based on Apache Spark to predict diabetes. Five machine learning classification methods were used like Decision Tree, Support Vector Machine, Logistic Regression Classifier, Naive Bayes, and Random Forest Classifier. Comparison between different algorithms was calculated using three measures, which are accuracy, recall, and precision. The experimental results proposed that LR achieved the highest percentage of accuracy, recall, and precision,82%, 92%, and 82%, respectively.