Analysis and Prediction of COVID-19 Pandemic in India

Abstract

In this paper, we have analysed the COVID-19 progression in India and the three most affected Indian states (viz. Maharashtra, Tamil Nadu and Andhra Pradesh) as of 29-Aug-20 and developed a prediction model to forecast the behaviour of COVID-19 spread in the future months. We used time series data for India and applied the Susceptible-Infective-Removed (SIR) model and the FbProphet model to predict the peak infectives and peak infective date for India and the three most affected states. In this paper, we further performed the comparative analysis of the prediction results from SIR and FbProphet models. From this study, we concluded that with the assumption that a total 5% of India's population might be infected by the pandemic, the countrywide spread is forecasted to reach its peak by the end of Nov-20. And till the time there is no vaccination, for the states that have already reached their peak and with festivals around the corner, there are high chances of resurgence in the number of cases if the social distancing and other control measures are not followed diligently in the coming months.