A Comparative Study of Naive Bayes, SVM, and Decision Tree Algorithms for Diabetes Detection Based on Health Datasets

Abstract

Diabetes is a chronic, progressive condition whose global prevalence continues to rise, creating substantial public health and economic burdens. Early diagnosis and timely intervention are critical to preventing severe complications and improving long-term patient outcomes. In recent years, artificial intelligence (AI) particularly machine learning (ML) has emerged as a powerful tool in medical diagnostics, offering capabilities in automated pattern recognition and disease classification. This study aims to evaluate and compare the predictive performance of three supervised ML algorithms such as Naïve Bayes, Support Vector Machine (SVM), and Decision Tree for classifying and predicting diabetes based on two primary physiological indicators: glucose level and blood pressure. The dataset employed was sourced from Kaggle, comprising 995 patient records containing relevant clinical attributes. The research methodology involved several stages, including data preprocessing to ensure quality and consistency, data partitioning into training and testing subsets using an 80:20 split ratio, model training, and performance evaluation. Each algorithm’s effectiveness was measured using accuracy, precision, recall, and F1-score metrics. The experimental findings demonstrate that the Decision Tree algorithm achieved the highest classification accuracy (94.47%), outperforming SVM and Naïve Bayes, both of which recorded 92.96% accuracy. Moreover, the Decision Tree exhibited balanced precision and recall values, underscoring its robustness in identifying both diabetic and non-diabetic cases with minimal misclassification. These outcomes indicate that the Decision Tree model provides an optimal balance between predictive accuracy and interpretability, making it particularly suitable for clinical decision-support applications.