Klasifikasi risiko diabetes menggunakan algoritma naive bayes berdasarkan dataset pima indians diabetes

Abstract

Diabetes mellitus is a non-communicable disease with a continuously rising prevalence, making early detection essential to prevent further complications. Conventional diagnostic methods that rely on laboratory examination are often constrained by time, cost, and limited access to healthcare facilities. This study aims to build and evaluate a diabetes risk classification model using the Gaussian Naive Bayes algorithm on the Pima Indians Diabetes dataset, consisting of 768 patient records with eight clinical variables. The research stages include data preprocessing to handle medically implausible zero values, splitting the data into training and testing sets using an 80:20 stratified scheme, model training, and evaluation using accuracy, precision, recall, F1-score, and Area Under Curve (AUC) metrics. The results show that the model achieved an accuracy of 70.13%, precision of 56.67%, recall of 62.96%, F1-score of 59.65%, and an AUC of 0.765. Glucose and BMI were identified as the most influential variables in the classification outcome, consistent with clinical understanding of type 2 diabetes risk factors. This study concludes that Gaussian Naive Bayes, despite its computational simplicity, remains a viable instrument for early diabetes risk screening, with a recommendation to apply class-balancing techniques in future research to improve model sensitivity.