APPLICATION OF MODEL-BASED CLUSTERING ALGORITHM TO COVID-19 VACCINE DATA

Abstract

In Covid-19 pandemic, countries have developed various policies to get over this period with minimum damage. These policies have been updated and are still being updated at each stage of the pandemic to maximize benefit to the society. Vaccination policies of countries have become crucial after vaccine was developed. Some inequalities such as opportunity of developed countries and inability of other countries to access vaccine and anti-vaccination are considerable hinders to prevent spread of the pandemic. We used Covid-19 data to cluster European Union Countries, Candidate Countries and Potential Candidate Countries. At the first stage of the study, optimum algorithm was determined with use of internal and stability validation indexes for clustering of countries. At the second stage of the study, model algorithm was applied and it was determined that there are 20 countries in the first cluster and 14 countries in the second cluster. In conclusion of the study, cluster-based variables analysis shows that deaths and positive rate are lower since vaccination rate is high no matter how high is the number of new cases and the reproduction rate.