<![CDATA[Diabetes mellitus is a chronic metabolic disease that requires early diagnosis to prevent serious complications. This study aims to develop a diabetes risk classification model using the Decision Tree algorithm, based on a Kaggle dataset consisting of 520 patient records with 17 clinical and demographic features. The evaluation results show that the proposed model performs well, achieving an accuracy of 95.19%, precision of 1.00, recall of 0.93, and an F1-score of 0.96 for the positive class. These results outperform six previous studies that employed algorithms such as C4.5, SVM, Adaboost, KNN, and Naïve Bayes, both in terms of accuracy and sensitivity in detecting diabetes cases. In addition to its strong performance metrics, the study highlights the model’s interpretability through a visualized decision tree, enabling healthcare professionals to better understand the classification logic. With broader attribute coverage and comprehensive performance evaluation, the model is considered effective and suitable for developing accurate and transparent decision support systems for early diabetes diagnosis.]]>
Copyrights © 2025
