Statistical Analysis, Modeling and Development of Probable Risk Index of COVID-19 Data

In the project, firstly, it is aimed to determine the behavior of the COVID-19 data with the help of descriptive statistics and to perform a cluster analysis for the COVID-19 data in order to determine the place of Turkey in the world in terms of the COVID-19 outbreak. Secondly, it is aimed to analyze the COVID-19 data with statistical models. Finally, it is aimed to create some indexes related to the COVID-19 epidemic by using the data of our country. Within the scope of the project, firstly, descriptive statistics and distributions that can model the data were determined for Turkey and some selected world countries representing each continent. According to the results of the cluster analysis, it was concluded that America and Europe have a more homogeneous structure in terms of the number of cases. Secondly, linear and nonlinear regression, breakpoint regression, logistic growth regression, ARIMA, power-law, polynomial time trend and loess regression models are discussed to explain the number of cases, recoveries and deaths. It has been determined that reliable predictions can be made for the short term with these models. Government Intervention Index (GRI), COVID Contagion Index (CCI) and Government Protection Index (GPI) have been established for our country. The effects of government interventions are presented according to GRI values. According to the CCI values, it has been determined that people in the 25-49 age range have the highest contagiousness, and people in the 20-24 and 65-74 age ranges have the lowest contagiousness. With the GPI value, it has been shown that the government's policy of protecting its citizens against the epidemic is successful. In addition, the website "http://covid19stats.ankara.edu.tr/" was prepared in order to disseminate the project results. As a result of the cluster analysis, Turkey was included in the groups with the highest number of cases, recoveries and deaths. In the light of the findings of the refractive regression model, it is suggested that more stringent measures should be taken due to the increase in the number of cases when the measures are removed. According to the State Intervention Index, one of the most influential measures on the index until the epidemic is completely brought under control; It has been concluded that curfews, interruption of face-to-face education and the continuation of travel restrictions are very important in preventing the epidemic. According to the age-related results of the COVID Contagion Index, it has been concluded that the infectivity values ​​of the age groups subject to restrictions are at the minimum level, and the infectivity coefficients are at the highest level due to the intense mobility of the working age groups.