PERBANDINGAN KINERJA ALGORITMA NAÏVE BAYES DAN DECISION TREE UNTUK PREDIKSI PENYAKIT BATU GINJAL

Abstract

Kidney stones consist of calcium crystals and are mineral deposits that can form in the urinary tract. The prevalence of kidney stones is estimated to be 21.11%, with the Standard Age Prevalence in men and women being 24.3% and 18.7%, respectively. This study involved participants with an average age of 52.15 years, and a higher prevalence of kidney stones was observed in women aged 40-50 years and individuals with a moderate socioeconomic status. Logistic regression results indicate that the likelihood of kidney stones is higher in individuals with diabetes, hypertension, fatty liver, and overweight. The Basic Health Research (Riskesdas) data for the year 2013 shows the prevalence of Kidney Failure and Kidney Stones in Indonesia. Therefore, a profound understanding of the factors causing kidney stone formation is crucial. This research was conducted to compare the performance of the naive Bayes and decision tree algorithm models. Based on the experimental results, the decision tree algorithm showed a higher accuracy rate of 72.50% and an AUC value of 0.740%, while naive Bayes had an accuracy rate of 68.57% and an AUC value of 0.697. These results support the conclusion that the decision tree has better predictive performance than naive Bayes.