CHANGING STATUS OF GLOBAL COVID-19 OUTBREAK IN THE WORLD AND IN TURKEY AND CLUSTERING ANALYSIS

Objective: In this study, it is aimed to provide a dynamic structure to the summary status and analysis results based on the current COVID-19 data of the countries based on changing status of global COVID-19 outbreak in the world and in Turkey; thus, to support fast and proactive decisions. In this scope, to define COVID-19 based on data, an online R-Shiny application is developed (https://elifkartal.shinyapps.io/covid19/). Material and Method: In this study, CRoss-Industry Standard Process for Data Mining - CRISP-DM is used as the study method. The changing situation of COVID-19 in global and national dimensions was evaluated. New variables are calculated such as Linear Change Rate (LCR), Exponential Growth Coefficient (EGC), and required days to double cases. Cluster analysis was performed by applying the k-Means data mining algorithm to the data reinforced with the new variables and similarities of countries were determined. The countries closest to the cluster average are accepted as cluster centers and the countries in the same cluster are ranked according to their distance from the cluster center. Results: One of the most important findings of the study is that the trends of LCR and EGC are the same. As such, it can be said that COVID-19 does not display an exponential behavior or can be controlled. With the developed application, the countries in which the cluster is located, regardless of their geographical location and dynamically according to time, the possible risk situations and similarities of the countries in the same cluster have been determined more precisely. Conclusion: With this study and the application developed; depending on changing status of global COVID-19 outbreak in the world and in Turkey, a dynamic structure has been given to the summary status and analysis results based on the current COVID-19 data of the countries, thus, it has been provided to support fast and proactive decisions.