Use of Machine Learning Techniques to Predict Diabetes at an Early Stage

Because of its ability to improve disease prediction, machine learning has taken a prominent place in healthcare services (HCS). Artificial intelligence and machine learning techniques have already been used in this area. In order to anticipate disease at an early stage, data mining techniques are commonly used. Diabetes has recently become a well-known public chronic condition all across the world. It is rapidly increasing as a result of improper lifestyles, increased consumption of junk food, and a lack of health awareness. Predictive analytics in healthcare is a difficult task, but it can ultimately assist practitioners in making timely decisions regarding a patient’s health and treatment based on huge data. For the purpose of predicting diabetes, seven of the most important machine learning classification techniques have been examined. As a result of a comparison of the multiple machine learning approaches utilized in this study, it has been determined which algorithm is best for prediction of diabetes. With an F1 score of 0,94, XGBoost outperformed other classifiers. To help doctors and practitioners anticipate diabetes earlier using machine learning approaches with more accuracy, this study was written. Models were shown to be more effective than existing work.