SIGNATURE WORK
CONFERENCE & EXHIBITION 2022

Diabetes is one of the most dangerous conditions of the 21st century, being the 8th leading cause of death in the world as of 2015 (Tao et al., 2015). Even though there are various genetics, lifestyle factors, and underlying conditions linked to the development of diabetes, research proves that type 2 diabetes can be preventable through a healthy diet and lifestyle changes (Hu, 2011).

This paper contributes to developing algorithms that can assist diabetes controlling technologies by comparing different machine learning models to identify individuals at high risk of developing diabetes and predict the glycemic response of certain lifestyle activities that would contribute to the prevention and treatment of diabetes. This paper uses three public diabetes datasets, including information on different genetics, lifestyle factors, and underlying conditions linked to diabetes and measured postprandial blood glucose levels. In addition, this work compares seven machine learning models, including logistic regression, random forest, support vector machine, and artificial neural network, by evaluating their accuracies, F-scores, and how suitable they would be in the medical field.