An Analytic Model for COVID-19 Cases in Iran and Its Neighbors Using Deep Learning and Time Series Methods

Abstract

     Since the pandemic of the coronavirus (COVID-19) in 2019, it has rapidly become a major global health concern. Various mutations and rapid spread of the virus, a lack of specific treatment, and limited hospital facilities highlight the importance of anticipation, risk analysis, and timely treatment. The use of mathematical models, artificial intelligence, and simulation methods are effective tools in predicting the spread and providing effective solutions to prevent virus transmission. Analysis and forecasting require an integrated model to cover different aspects of the problem and use different methods to obtain appropriate results.    In this research, a proposed model for analysis and prediction of COVID-19 cases in Iran and neighboring countries is presented. The performance of mathematical and deep learning models in the proposed model has been evaluated using data from Johns Hopkins University from January 29, 2020, to April 30, 2021. Evaluation of the predictive outcomes of daily cases was performed using RMSE criteria. Then, the effect of the trend of cases in neighboring countries of Iran on the rate of new cases in this country has been studied. These models can help governments predict the number of infections to provide the necessary solutions and prevent a new wave of the virus.