The Application Of Particle Swarm Optimization Using Naive Bayes Method For Predicting Heart Disease

Abstract

Heart disease is a disease that has many sufferers and is one of the deadliest diseases in the world. Based on the 2018 Basic Health Research Report, the average prevalence of heart disease in the country (Indonesia) was 1.5% that year. It is recorded that 11 provinces have a prevalence of heart disease above the national average. North Kalimantan has the highest prevalence of heart disease in Indonesia at 2.2%, Yogyakarta and Gorontalo at 2%, East Kalimantan, DKI Jakarta and Central Sulawesi at 1.9%, North Sulawesi at 1.8%, Aceh, West Sumatra, West Java, Central Java by 1.6%, and East Nusa Tenggara by 0.7%. The cause of the increase in death rates every year is due to lack of access to find information about heart attack disease. From this problem, researchers want to develop technology in the health sector, especially using a data mining classification algorithm, namely the Naive Bayes algorithm. In the previous research, namely the journal entitled A Clinical support system for Prediction of Heart Disease using Machine Learning Techniques with the same dataset applying the Naive Bayes algorithm without using feature selection, the accuracy value was 82.17% and the split data accuracy value was 84.28%. In this study, the researcher applyed the Naive Bayes algorithm using the Particle Swarm Optimization feature selection, the accuracy value is 84.16% and the split data accuracy is 85.12%. so it can be concluded that the Particle Swarm Optimization selection feature can be used to optimize the accuracy value of the Naive Bayes algorithm.

Keywords—Heart Disease, Data Mining, Naïve Bayes and Particle Swarm Optimization