Analisis Perbandingan Klasifikasi Penyakit Jantung Menggunakan Algoritma Nai ̈ve Bayes dan Algoritma Logistic Regression

Abstract

Heart disease is a condition where parts of the heart are damaged. Thus, early detection is needed. One of them is by doing data mining classification using the Naïve Bayes and Logistic Regression algorithms. This research will compare Naïve Bayes and Logistic Regression algorithms through the training percentage split and k-fold cross validation methods to get the best classification results in detecting heart disease by calculating the average value of precision, recall, and accuracy. The Naïve Bayes algorithm with the training percentage split method produces average values for precision, recall, and accuracy of 83%, 82.5% and 81%, while the Naïve Bayes algorithm with k-fold cross validation provides average values for precision, recall, and accuracy of 83.5%, 85.5% and 83%. Logistic Regression algorithm with percentage split training method produces average values for precision, recall, and accuracy of 73.5%, 73.5% and 73%, while Logistic Regression algorithm using k-fold cross validation produces average values for precision, recall, and accuracy of 84%, 83.5% and 84%. This shows that the Naïve Bayes algorithm using percentage split is better than Logistic Regression, but when using the k-fold cross validation method, the Logistic Regression algorithm has a significant increase compared to Naïve Bayes. So that to classify heart disease is better with the Logistic Regression Algorithm with the k-fold cross validation method.