Early-Stage Diabetes Risk Prediction Using Machine Learning Techniques

Diabetes is a chronic and metabolic disease caused by an increment in glucose level in the blood, which leads overtime to major serious damage to the heart and kidneys. Millions of people are affected by diabetes in the world. The main aim of this study is the prediction of the risk of diabetes at an early stage. The data set contains information on the symptoms of newly diagnosed diabetic patients, those who are at risk of developing diabetes. The data was collected through direct questionnaire's administer to patients at the hospital in Bangladesh. In this work, 10 supervised machine learning algorithms such as Random Forest (RF), Naive Bayes (NB), etc. are implemented by taking accuracy, precision, F1 score etc. as evaluation criteria. This diabetic risk prediction model has a higher accuracy of 99.12% with Random Forest and Gradient boosting than other models.