PEMODELAN KLASIFIKASI PREDIKSI DIABETES MENGGUNAKAN PENDEKATAN ALGORITMA NAÏVE BAYES

Diabetes Mellitus is a global chronic disease according to WHO whose prevalence continues to increase massively, making early detection and risk prediction crucial for effective prevention and treatment. The dataset used in this study consists of medical and demographic parameters of a number of patients who have been diagnosed with diabetes as well as those who have not. The Naïve Bayes method was chosen for its ability to handle complex data with the assumption that each feature is independent of the other. Data pre-processing steps including normalization, removal of missing values, and feature selection were performed prior to the application of the classification algorithm. Model
evaluation is performed using performance metrics such as accuracy, Classification Error, AUC, precision, recall, F-measure, and sensitivity. The experimental results show that the Naïve Bayes model is able to provide a fairly good prediction of diabetes status, with an average accuracy of 95.5%. These
findings demonstrate the potential of using the Naïve Bayes algorithm approach in the development of diabetes prediction systems that can be used in clinical practice to support more timely and accurate decision making.
Keywords: Diabetes Mellitus, Classification, Naïve Bayes algorithm, Prediction