When to Relax Social Distancing Measures? An ARIMA Based Forecasting Study

TL;DR

This study uses ARIMA models on positive rate data to forecast COVID-19 transmission and inform decisions on relaxing or tightening social distancing measures across different countries.

Contribution

It introduces ARIMA-based forecasting of positive rates to guide social distancing policy decisions during the pandemic.

Findings

ARIMA models effectively fit positive rate data

Forecasts suggest optimal timing for relaxing measures

Some countries need to tighten restrictions based on forecasts

Abstract

The spread of the novel coronavirus across various countries is wide and rapid. The number of confirmed cases and the reproduction number are some of the epidemiological parameters utilized in scientific studies for the analysis and prediction of the viral transmission. The positive rate, an indicator on the extent of testing the population, aids in understanding the severity of the infection in a given geographic location. The positive rate for selected countries has been considered in this study to construct ARIMA based statistical models. The goodness of fit of the models are verified by the investigation of residuals, Box-Luang test and the forecast error values. The positive rates forecasted by the ARIMA models are utilized to investigate the scope for implementation of relaxations in social distancing measures in some countries and the necessity to tighten the rules further in some other countries.