Diabetes is a chronic metabolic disorder that affects millions of people worldwide and is considered one of the most serious long-term diseases due to its complications. Early diagnosis and prediction of diabetes are crucial for preventing severe health outcomes. In recent years, Machine Learning (ML) techniques have emerged as powerful tools in healthcare, particularly in disease prediction and diagnosis. The objective of this study is to develop an accurate predictive model for assessing the likelihood of diabetes in patients. The Pima Indian Diabetes Database (PIDD) from the UCI Machine Learning Repository is utilized for experimentation. Three ML classification algorithms—J48, K-Nearest Neighbor (K-NN), and Naïve Bayes (NB)—are employed to detect diabetes at an early stage. The models are evaluated using performance metrics such as precision, accuracy, recall, and F-measure. Results indicate that the J48 algorithm outperforms the others, achieving an accuracy of 82%. In comparison, Naïve Bayes achieved 76% accuracy, while K-NN recorded the lowest accuracy at 74% using the same dataset. Future work aims to develop hybrid or ensemble models to enhance predictive accuracy and reliability.
