Implementing Machine Learning Approach for the Prediction of Cumulative COVID-19 Cases in Nigeria

Abdallahi J

Department of Civil Engineering, Baze University, Abuja, Nigeria

^*Correspondance to: Jazuli Abdallahi 

 PDF  Full Text Research Article | Open Access

Abstract:

Since December 2019, Coronavirus Diseases (COVID-19) has been a major topic of interest to the researchers, authorities and health officials due to its devastating nature and risk associated with its contraction and spread. Developing countries including Africa are at risks of getting more infected due to less awareness and ill-equipped health facilities. For such countries, accurate prediction of the COVID-19 cases could help the policy makers in decision making and proper utilization of the available resources. Machine Learning (ML) approaches are excellent tools for making predictions especially for an uncertain and stochastic phenomenon such as COVID-19. Therefore, Artificial Neural Network (ANN) as ML tool was employed to predict cumulative COVID-19 cases in Nigeria. In this way, data including daily and cumulative cases as well as daily and cumulative mortality rate were collected from February 2020 till December 2021. Five different models were developed using varied input combinations. Determination Coefficient (DC) and Root Mean Square Error (RMSE) were employed as performance metrics. The results showed all the developed models produced good prediction skill but M4 with DC=0.9996 and RMSE=0.0017 in the validation step led to better performance. The general results of this study emphasize the importance of using ANN for the prediction of COVID-19 pandemic in Nigeria.

Keywords:

COVID-19; Machine learning; Pandemic; Nigeria; Africa

Cite the Article:

Abdallahi J. Implementing Machine Learning Approach for the Prediction of Cumulative COVID-19 Cases in Nigeria. Clin Case Rep Int. 2022; 6: 1424.