A Predictive Model for Confirmed Cases of COVID-19 in Nigeria

Abstract

Background and Objectives: COVID-19 pandemic globally remains a major problem affecting every aspect of human endeavours with Nigeria, not an exception. The number of confirmed cases of COVID-19 in Nigeria as at 8th May, 2020 was 3,912 this study builds an ARIMA model for forecasting the confirmed cases of COVID-19 in Nigeria based on Box-Jenkins methodology. Materials and Method: Data on confirmed cases of COVID-19 in Nigeria were obtained from the Nigeria Centre for Disease Control (NCDC). The stationarity of the data was determined using the Augmented Dickey-Fuller (ADF) test and stationarity was achieved after taking the natural log transformation and first differencing. Based on Autocorrelation Function (ACF) and Partial Autocorrelation Function (PACF) plots, three ARIMA models were identified [ARIMA (2,1,0), ARIMA (0,1,2) and ARIMA (2,1,2)]. The fitness performance of these models was compared using R2 and normalized Bayesian Information Criteria while the forecasting accuracy was compared using Root Mean Square Error (RMSE). Results: Results showed that among the ARIMA models, the ARIMA (2,1,0) outperformed other proposed models both in terms of fitness and forecasting accuracy and hence the ARIMA (2,1,0) was recommended for forecasting confirmed cases of COVID-19 in Nigeria. Hence, the ARIMA (2,1,0) model was then used to project the confirmed cases of COVID-19 in Nigeria for the next two weeks (9/05/2020 to 2/05/2020). The forecasts showed an upsurge in the confirmed cases of COVID-19 in Nigeria if the current relaxation of the lockdown continues. Conclusion: This finding has implication for the need for the Nigeria government to ensure that COVID-19 preventive measures such as social distancing, use of face mask, regular hand washing with soap and water or alcohol-based sanitizer as well as the restriction of movement are sustained to possibly hurt these projected increase in the confirmed cases of COVID-19 in Nigeria in order to prevent the collapse of health systems.