Clustering of European Countries in terms of Healthcare Indicators

Abstract

Health is always considered as one of the most important issues related to human being. Due to this importance, governments should primarily provide the best healthcare services to their citizens. Some indicators can show the quality of healthcare services in the country. However, one country can have a higher value of one indicator and can have a lower value of another. Thus, countries can be categorized in terms of quality of healthcare services. Clustering is a useful tool for comparing countries and defining the similar countries in terms of healthcare services. In this study, 28 European Union (EU) countries were evaluated on 14 health factors and the number of clusters was determined by the generally accepted rule of thumb. To cluster countries, k-means clustering method is run in WEKA software for two cluster numbers and four different initial solution approaches. The resulting clusters were evaluated according to the Spearman rank correlation coefficient using the order of the GDP per capita values of the countries in each cluster. It seems using four clusters with Canopy initial solution approach is the most appropriate way of clustering.

Keywords

• Data mining,Clustering,Healthcare indicators,EU countries

References

1. F. Ersöz Türkiye ile OECD ülkelerinin sağlık düzeyleri ve sağlık harcamalarının analizi, İstatistikçiler Dergisi, 2 (2008) 95–104.
2. F. Lorcu, B. A. Bolat, A. Atakisi, Examining Turkey and Member States of European Union in Terms of Health Perspectives of Millennium Development Goals, Quality & Quantity, 46 (2012) 959-978. DOI: 10.1007/s11135-011-9648-1
3. N. Alptekin, Comparison of Turkey and European Union Countries’ Health Indicators by Using Fuzzy Clustering Analysis, International Journal of Business and Social Research, 4 (2014) 68-74. DOI: 10.18533/ijbsr.v4i10.607
4. K. E. Lewandowski, S. H. Sperry, B. M. Cohen, D. Öngür, Cognitive variability in psychotic disorders: a cross-diagnostic cluster analysis, Psychological Medicine, 44 (2014) 3239-3248. DOI: 10.1017/S0033291714000774
5. D. K. Moser, K. S. Lee, J. R. Wu, G. Mudd-Martin, T. Jaarsma, T. Y. Huang, X.Z. Fan, A. Strömberg, T. A. Lennie, B. Riegel, Identification of symptom clusters among patients with heart failure: An international observational study, International Journal of Nursing Studies 51 (2014) 1366-1372. DOI: 10.1016/j.ijnurstu.2014.02.004
6. C. H. Olson, S. Dey, V. Kumar, K. A. Monsend, B. L. Westra, Clustering of elderly patient subgroups to identify medication-related readmission risks, International Journal of Medical Informatics 85 (2016) 43-52. DOI: 10.1016/j.ijmedinf.2015.10.004
7. S. Tsumoto, S. Hirano, H. Iwata, Mining Schedule of Nursing Care based on Dual-Clustering, Procedia Computer Science, 55 (2015) 1203-1212. DOI: 10.1016/j.procs.2015.07.125
8. J. MacQueen, Proc. Fifth Berkeley Symposium on Mathematical Statistics and Probability, 21 June – 18 July, 1965 and 27 December, 1965 - 7 January, 1966 Berkeley-USA.

9. T. M. Kodinariya, P. R. Makwana, Review on determining number of Cluster in K-Means Clustering, International Journal of Advance Research in Computer Science and Management Studies, 6 (2013) 90-95.
10. World Bank (26.03.2016). World Development Indicators. Link: http://data.worldbank.org/data-catalog/world-development-indicators Access data: 26.03.2016.