A New Deep Learning Model for Early Stage of Diabetes Disease Prediction

Diabetes mellitus is a chronic disease caused by several factors, including age, obesity, physical inactivity, inheritance, and unhealthy nutrition. Therefore, early stage detection of chronic diseases is of utmost importance for individuals. In this light, intelligent approaches are needed to predict the risk of diabetes mellitus. In this study, a new deep learning model was proposed to estimate the risk of diabetes disease. The performance of the new model was compared to the popular machine learning algorithms. In order to improve model performance, hyperparameter optimization was handled by the Grid Search algorithm. The proposed model achieved 98.71% accuracy, 98.74% precision, 98.71% recall, and 98.71% F1-score performance on the UCI diabetes data set. The proposed model outperformed compared machine learning models.