Optimization of the Naïve Bayes Algorithm Using Particle Swarm Optimization (PSO) for Predicting Heart Disease Symptoms

Muharni, Sita
Andriyanto, Sigit
Supardi, Supardi

Prisma Sains: Jurnal Pengkajian Ilmu dan Pembelajaran Matematika dan IPA IKIP Mataram
Vol 13, No 3: July 2025

Heart disease is one of the leading causes of mortality in Indonesia, yet early detection remains a challenge due to limitations in data and suboptimal classification methods. This study aims to improve the accuracy of heart disease prediction by integrating the Naïve Bayes algorithm with Particle Swarm Optimization (PSO) for feature selection. A dataset of 303 patient records was processed using RapidMiner across three configurations: Naïve Bayes with split validation (80:20), Naïve Bayes with 10-fold cross-validation, and Naïve Bayes with PSO-based feature selection. The results showed that incorporating PSO increased accuracy from 87.60% to 89.26%, along with improvements in precision and recall, while maintaining a high AUC value (0.933). These findings demonstrate that PSO effectively identifies the most relevant features and enhances the performance of heart disease prediction models. The study also underscores the importance of validation methods and model interpretability in the application of artificial intelligence in healthcare.