Analyzing covid-19 outbreak for turkey and eight country with curve estimation models, box-jenkins (arima), brown linear exponential smoothing method, autoregressive distributed lag (ardl) and seir models

Abstract

Aim: This study is conducted to inform communities and governments about the spread of the COVID-19 pandemic in selected countries: Turkey, Germany, the United Kingdom, France, Italy, Russian, Canada, and Japan. Materials and Methods: For this purpose, the numbers of the COVID-19 epidemic after the 100th case up to 7/19/2020 for selected countries have been estimated by using Curve Estimation Models, Box-Jenkins (ARIMA), Brown Linear Exponential Smoothing Method, Autoregressive Distributed Lag (ARDL) and SEIR Models. Results: In the evaluations of the ARDL and SEIR models established, it is determined that France and Italy have high pandemic growth rates; while Canada has a low pandemic growth rate. It has also observed that the turning point of the pandemic occurred on the 72nd day. If there is no change in the outbreak and governments continue with the same strategies, it is predicted that the epidemic will begin again in early October 2020 (from September 21 to November 10) and will be effective for an average of 155 days (between 145 and 168 days). It is seen that the observed and predicted daily cumulative new cases are consistent. Conclusion: As a result, it can be said that the models used in this study wellcharacterized outbreak of the COVID-19 in the eight major Western countries and Turkey.

Keywords

• covid-19
• curve estimation models
• arima
• exponential smoothing methods
• ardl
• seir
• box-jenkins models