A Time Series Model to Forecast Covid-19 Infection rate in Nigeria Using Box-Jenkins Method

Coronavirus declared as a global pandemic by WHO has emerged the most aggressive disease negatively affecting more than 90% countries of world. Nigeria, one populated in Africa is not an exception. This study focuses on analyzing intrinsic patterns COVID-19 spread Nigeria using Box-Jenkins procedure. Data daily confirmed cases was retrieved from Centre for Disease Control (NCDC) official website February 27, 2020 to October 31, identify series components, estimate parameters, develop appropriate stochastic predictive model and use forecast future trend deadly virus. The Autoregressive Integrated Moving Average (ARIMA) order (0,1,1) identified suitable based analysis autocorrelation (ACF), partial functions (PACF) Akaike Information Correction (AICc) value. R software version 4.0.3 used analyze which moothen 8-point moving average extract irregular component wellas differencing step further obtain stationary series. We performed Augmented Dickey-Fuller Unit root test, parameter estimation Ljung-Box test check proposed model’s conformity univariate process. A 85 – day (1st Oct., 24th Jan., 2020)forecast shows gradual decline successive number infection indicating effectiveness intervention strategies employed Task Force contain concerned authorities can apply forecasted make informed decisions measures be put place reduce diffusion virus into country.