Forecasting of COVID-19 Cases Using SARIMA Model in India

Authors

  • Aashish Jain, Ram Chandra Singh, Rohit Khokher, Rahul Kumar

Abstract

On January 30, 2020, the World Health Organization (WHO) declared a global health
emergency for coronavirus disease 2019 (COVID-19). The disease spread around the world and on March
11, 2020, it was declared a global pandemic by the WHO. India reported the very first case of COVID-19 on
January 30, 2020, and a complete lockdown on the country was announced on March 25, 2020, after WHO
announced that it was a global pandemic. Since then, India has reported 586K confirmed cases, of which
348K have been cured, 17.4K died and 220K is still active as of June 30, 2020. In this work, a predictive
model for the seasonal autoregressive integrated moving average (SARIMA or seasonal ARIMA) is
proposed to predict the total number of confirmed, recovered, and deceased cases due to the COVID-19
virus in India within the next 30 days from a selected date. The SARIMA model uses the Box-Jenkins
model, a forecasting method using regression studies. Data for this study has been taken from government
websites from January 30, 2020, to June 30, 2020. Using the Tkinter library in Python 3.8, a graphical user
interface (GUI) is also developed to make this prediction model user-friendly. The accuracy in predicting
COVID-19 cases in this study is 99.7% in confirmed cases, 99.15% in recovered cases, and 99.08% in
deceased cases.

Published

2020-03-25

Issue

Section

Articles