Clustering Countries by K-Means Method According to Causes of Death

Abstract

Causes of death are one of the criteria used to assess countries' health systems and determine their Human Development Levels. Countries are developing health policies based on the causes of death. While mortality rates and causes of death are accepted as development indicators for countries by the United Nations, improvement of public health is considered as a global target. According to the Institute for Health Metrics and Evaluation, 54.15 million deaths occurred in 2015, 71% of which were caused by non-communicable diseases, 20% were caused by communicable diseases, neonatal and nutritional diseases, and the remaining 9% were caused by injuries. In this study, it is aimed to group the countries by considering various causes of death of people in different countries and to investigate whether there is a relationship between the causes of death and the Human Development Level of the countries. In the analysis; 2015 data of 168 countries and 28 different variables showing the causes of death of these countries were used. K-means method was used to group the countries according to causes of death and 4 different models were established by making use of World Health Organization's classification of illness, injury and causes of death. After the cluster analysis, in which clusters the countries are located according to Human Development Level were examined. It is also investigated that whether there is a relationship between the causes of death and the Human Development Level of the countries.

Keywords

• cause-related death
• cluster analysis
• k-means