Performance of Machine Learning Algorithms (Decision Tree an‎d TreeNet) in Prediction an‎d Classification of Type 2 Diabetes

11/22/2025 12:00:00 AM

In recent years, machine learning has emerged as one of the effective tools for predicting an‎d diagnosing diseases in the medical field. The aim of this study is to examine an‎d compare the performance of two machine learning algorithms—Decision Tree an‎d Gradient Boosted Tree (Tree Net model)—in predicting the incidence of type 2 diabetes. Due to limited access to real data, simulated data were generated based on the characteristics of the Pima Indians Diabetes dataset. These data included various medical features such as the number of pregnancies, glucose level, blood pressure, body mass index, an‎d patients’ age. After data preprocessing, both models were eva‎luated using various metrics such as accuracy, sensitivity, specificity, an‎d F1-score. The results showed that the Tree Net model had nearly equal predictive capability compared to the Decision Tree according to the eva‎luation metrics. Moreover, graphical analysis of the models’ performance confirmed the superiority of the Tree Net model. The findings of this study confirm the applicability of advanced machine learning algorithms in disease prediction an‎d pave the way for further research to improve clinical decision support systems.

يادگيري ماشين، الگوريتم درخت تصميم، الگوريتم TreeNet، ديابت نوع 2. , آنتروپي , شاخص جيني , ماتريس در هم ريختگي

Machine Learning, Decision Tree Algorithm, TreeNet Algorithm, Type 2 Diabetes , entropy , Gini Index , Confusion Matrix

Morteza rozbahani

Dr. Rahman Farnoush