Determining the efficiency of data analysis systems in predicting COVID-19 infected cases

After the outbreak of the novel coronavirus disease (2019) (COVID-19), a lot of people have been affected around the world. Due to the large number of affected patients in the world, the global health care system has been disrupted and nearly all hospitals around the world has faced a shortage of bed spaces. As a consequence. being able of prediction of the number of COVID-19 cases is extremely important for taking appropriate decision for management of the affected patients. An accurate prediction of the number of COVID-19 cases Can be obtained using the historical data of reported cases as well as some other data affecting the virus outbreak. However, most of the literature has used only historical data to provide a method of predicting COVID-19 cases and has neglected other influential factors. This has led to inaccurate estimates of the number of infected cases with COVID-19. Thus, the present study tries to provide a more accurate estimation of the number of COVID-19 cases by considering both historical data and other effective factors on the virus. For this purpose, data analysis including the development of a network-based neural algorithm [i.e., nonlinear autonomous exogenous input (MARX)] can be adopted. To examine the viability of this algorithm, experiments were conducted using data collected for the number of COVID-19 cases in the five most affected countries on each continent. Our method led to a more accurate prediction than those obtained by the existing methods. Moreover, we performed experiments to extend our method to predict the number of COVID-19 cases in the future during a period between August 2020 and September 2020. Such predictions can be utilized by the government or people in the affected countries to take precautionary measures against the pandemic.