Abstract

The first few cases of the novel coronavirus outbreak can be traced back to those that occurred in the Chinese city of Wuhan in December, 2019. On 30th January, 2020, the outbreak was declared as a Public Health Emergency of International Concern and later, on 11th March, 2020, it was declared to be a pandemic. As of 24thMay, 2020, it spreaded over 188 countries and territories and the total number of cases stood out at over 5.4 million, with over 346,427 deaths. In this paper, we restrict our attention to the outbreak in India. The first positive case was reported on 30th January, 2020 as per records. Number of confirmed cases (in India) stands at 138,535 with 4023 deaths as of 24th May. The main focus of our paper is to propose a Hybrid ARIMA (Auto Regressive Integrated Moving Average) model with error remodeling using Fourier Analysis performed on the number of daily new cases. Our aim is to show how the hybrid model generates better short-term forecasts compared to single ARIMA model and also how the modeling on a data set of the first phase of the lockdown generates more accuracy in the forecasting. These short-term forecasts for the number of daily new cases can guide us and throw some light on the growth pattern of COVID-19, thus guiding the government to make arrangements accordingly.

Author: Mehuli Paul and Meghanto Majumder

Received on: May, 2020

Accepted on: March, 2021