Diabetes Prediction Through Machine Learning and Ontology

Diabetes is a chronic metabolic disorder with a growing global impact on healthcare. Early detection and timely intervention are critical in preventing severe complications and improving patient outcomes. Recently, machine learning techniques and data framework-based approaches have played an important role in medical science by creating automated systems to identify diabetic patients. This paper reviews and compares popular machine learning methods and data framework-based classification techniques. The algorithms studied include Support Vector Machines (SVM), K-Nearest Neighbors (KNN), Artificial Neural Networks (ANN), Naive Bayes, Logistic Regression, and Decision Trees. The performance was measured using metrics like Recall, Accuracy, Precision, and F-Measure from the confusion matrix. This study evaluates six machine learning models on 768 samples from Kaggle's Pima Indian Diabetes Dataset. The results indicate that Ontology-based classification and SVM achieved the highest accuracy, making them highly effective for diabetes prediction.