Modelling and Prediction of Covid-19 Epidemic in Turkey Comparing with USA and China

Authors : Engin Ergül, Halil Ibrahim Kurt, Murat Oduncuoğlu, Necip Fazıl Yilmaz

Pages : 111-126

Doi:10.30931/jetas.790465

View : 15 | Download : 10

Publication Date : 2021-08-31

Article Type : Research

Abstract :The aim of the study is to research and compare the influences of the confirmed cases, test number and time range on the death and recovery rates in the United State of America, China, and Turkey, and to find out the effect of the epidemic in the near future of Turkey. The modelling and prediction of effects of the day, case and test numbers of COVID-19 infection in the USA, China and Turkey are carried out using the artificial neural network approach (ANN). The system are trained and tested with the different numbers of neurons, hidden layers and activation functions to increase the reliability and accuracy of model. The proposed models have a high R2 value for China and Turkey. We can say according to the results that the measures taken by the USA are inadequate. The formulation is applied to predict the effect of Covid-19 infection in Turkey. The test number that is an important factor in detecting the cases should be increased. The results show a good fit between the observed data and those obtained by the ANN model. If the precautions are strictly followed, the case number will be decreased significantly after 160 days for Turkey according to result of the proposed model but due to the uncontrolled variables, this time may result in between 200 and 250 days.
Keywords : Covid-19, Modelling, Prediction, USA, China, Turkey