A COMPARATIVE ANALYSIS OF MACHINE LEARNING ALGORITHMS FOR PREDICTING DIABETES MELLITUS

Diabetes mellitus continues to pose a significant global health challenge, with increasing prevalence across both high and low income countries. The inability to detect and treat diabetes effectively at its early stages has fueled the need for reliable predictive models. This research addresses the problem of determining the most accurate machine learning algorithm for predicting diabetes. The aim is to perform a comparative analysis of four machine learning algorithms, Random Forest (RF), Extreme Gradient Boost Classifier (XGBC), Support Vector Machine (SVM), and Naïve Bayes Classifier (NBC) to assess their predictive accuracy. The study uses diabetes datasets from the UCI Machine Learning Repository, applying feature selection, normalization, and statistical evaluation metrics like accuracy, precision, recall, and F1-Score to compare these models. The Results showed that the Random Forest algorithm outperforms the others, achieving 99% accuracy, followed by XGBoost at 93%, while SVM and Naive Bayes performed less effectively with accuracies of 90% and 83%, respectively. This project provides valuable insights into the effectiveness of different machine learning techniques for diabetes prediction, guiding healthcare professionals in model selection for early diagnosis and management.