To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
A Machine Learning Univariate Time Series Model for Forecasting COVID-19 Confirmed Cases : A Pilot Study in Botswana |
|
Author |
Ofaletse Mphale, Ezekiel U Okike, and Neo Rafifing |
| Citation |
Vol. 24 No. 12 pp. 101-110 |
|
Abstract |
The recent outbreak of corona virus (COVID-19) infectious disease had made its forecasting critical cornerstones in most scientific studies. This study adopts a machine learning based time series model - Auto Regressive Integrated Moving Average (ARIMA) model to forecast COVID-19 confirmed cases in Botswana in 60 days period. Findings of the study confirmed COVID-19 cases steadily rising with trend of random fluctuations and non-constant variance. This trend could be effectively described using an additive model in Seasonal Trend Decomposition procedure using Loess (STL). In selecting the best fit ARIMA model a Grid Search Algorithm (GSA) was used. The Akaike Information Criterion (AIC) metrics was used to derive scores of the different fit ARIMA models. In this study ARIMA model (5, 1, 1) with corresponding AIC score of 3885.091 was nominated. This model depicted the least value of the AIC measure. The forecasts results obtained from the study proved that ARIMA model could efficiently provide reliable estimates of forecasts that could be used to guide on understanding of future spread of COVID-19 confirmed cases. Findings of the study were useful in raising social awareness to disease monitoring institutions and government regulatory bodies where it could be used to support strategic health decisions and initiate policy improvement procedures for better management of the COVID-19 pandemic.. |
|
Keywords |
COVID-19, Corona virus, ARIMA, Box-Jenkins, Time series, Machine learning, ACF, PACF, AIC |
|
URL |
