Clustering of Risk Factors for Coronary Heart Disease Using The k-prototypes Algorithm

Coronary heart disease is a disease that plays a major role as a cause of death in the world. Data from the World Health Organization (WHO) states that an average of 17.9 million people die each year, which is around 31 percent of the total deaths globally. The Institute for Health Metrics and Evaluation (IHME) also reports that 14.4 percent of the causes of death in Indonesia are coronary heart disease. The magnitude of the influence of coronary heart disease on death results in the need for earlier knowledge about the characteristics of each patient with coronary heart disease. Each patient has different behavior and characteristics towards the cause of coronary heart disease. To find out each characteristic for people with coronary heart disease, need a data to be analyzed. Currently, the amount of data is getting bigger and many are of mixed type, including data on coronary heart disease patients. . The aim of this research is to classify coronary heart disease patient data using the k-prototypes algorithm. The best cluster result is k=2, with a silhouette coefficient value of 0.808. Cluster 1 of 493 patients, most of whom were male, had smoking habits with lower averages for the variables age, glucose, body mass index (BMI), diastolic, and heart rate, and had no history of hypertension, blood pressure medication (BPMeds), stroke or. For cluster 2, only 18 female patients had a history of hypertension, diabetes, and all of the numerical variables had a higher average than cluster 1.