Perbandingan Algoritma ID3, Naive Bayes, SVM Berbasis PSO untuk Prediksi Serangan Jantung

Prayogi, M. Bagus
Irawan, Indra
Fajar, Yahya Ibnu

MDP Student Conference
Vol 3 No 1 (2024): The 3rd MDP Student Conference 2024

This research aims to evaluate the precision of three primary predictive algorithms—ID3, naïve bayes, and SVM (SVM)—optimized using the particle swarm optimization (PSO) algorithm for detecting and predicting heart attacks. The methodology involves comparing these algorithms as tools for categorizing data into relevant groups and optimizing them using PSO to enhance prediction accuracy. Data from kaggle and uci repositories are managed using RapidMiner. The study reveals accuracy results: the ID3 algorithm achieves 75.20% accuracy with AUC 0.735, post-PSO optimization increases accuracy to 80.49% with AUC 0.815. The naïve bayes algorithm attains 81.52% accuracy with AUC 0.890, post-PSO optimization enhances accuracy to 83.94% with AUC 0.901. The SVM (SVM) algorithm records 82.13% accuracy with AUC 0.895, post-PSO optimization boosts accuracy to 84.83% with AUC 0.900.