Examination of the number of cases in Turkey COVID-19 using Markov chains model and model SEIARD

It was stated that after the Covid-19 epidemic emerged in China in December 2019, it spread to the whole world within three months. The first case in Turkey was seen on March 11, and on April 1, 2020, Covid-19 spread all over Turkey. In this study, it was examined how the course of the Covid-19 epidemic might change. First of all, the Covid-19 data announced by the Ministry of Health of Turkey until 31 December 2020 were used in the models created, and the change in the course of the disease was examined.Then, by defining the disease processes as states, transition possibilities between states were created.In addition, the progression, recovery, and extent of the dise ase wereexamined. The Markov chain model, which is one of the most useful models, is used to describe states and predict transition probabilities between states (Liu, 2020). The Markov chain models used are models that help to measure and produce reasonable and intuitive predictions. In addition, transition matrices were created by using Covid-19 data, a statistical method, Case Fatal Rate (CFR). These transition matrices are used in the generated models. Firstly, a 3-state Markov chain model was constructed, in which the transition between "Infected", "Healed" and "Death" states was observed. The Covid-19 data published by the Turkish Ministry of Health until December 31, 2019, were used in the model created, and case fatality rates were found. The data obtained in the 3-state Markov chain model showed that the epidemic can spread very quickly, the number of infected can reach serious dimensions and the case fatality rate may reach a situation that cannot be prevented as time progresses. By adding new states to this model, 5 more complex 5 states are observed, where transitions are observed between "Non-infected", "Infected" (isolated at home or hospital), "Infected" (Inpatient), "Intensive care" and "Death". Stated Markov chain model was created. Since the 5-state Markov chain model includes both temporal and recurrent states, it is a model that can be used to obtain long-term mortality probabilities and expected death times. With this model, the mean time to death of an infected person from a condition to the moment of death was obtained (Romeu, 2020). Then, a mathematical model, the SEIARD mathematical model, was used to investigate the Covid-19 outbreak (Avila et al., 2020). This model has been analyzed in detail, and its application has been made using Covid-19 data. With this model, the basic reproduction number (RO) was calculated and the daily infected, death and recovery rates were estimated. By minimizing errors, the parameters of the SEIARD mathematical model were created according to the reported data, and the development of the epidemic after 31 December 2021 was predicted. For this, the differential equation model is used. The model consists of the dynamics of six subpopulations. The change in the case fatality rate was examined with the established SEIARD mathematical model with 6 populations. The data obtained with the SEIARD mathematical model showed that the epidemic decreased over time and tended to disappear. With the help of Matlab program, models were created and results and graphics were obtained and examined.