View article

Slightly earlier illness forecast assists patients in preventing possibly life health concerns before it is too late, hence improving quality of health care. When coupled with Data Mining techniques, Machine Learning, a sub-set of Artificial Intelligence, has the potential to be effective in the area of predictions. This study aims to aid medical practitioners in early and accurate Type 2 diabetes detection and diagnosis. Based on eight clinical measurements from the widely used Pima dataset (its missing values were carefully handled), we use supervised machine learning techniques to improve the accuracy of diabetes mellitus prediction. Three distinct machine learning approaches were used: Decision Tree, Random Forest, and Gradient Boosting algorithm, with performance analysis. In addition to accuracy, F1-score, and Matthews Correlation Coefficient, the evaluation was based on a variety of performance factors. The …