Outbreak Prediction of COVID-19 for Dense and Populated Countries Using Machine Learning

Area of Science:

• Epidemiology
• Public Health
• Machine Learning

Background:

• The COVID-19 pandemic continues to significantly impact global health and strain public health resources.
• Managing the escalating number of COVID-19 cases presents substantial challenges for governing bodies worldwide.

Purpose of the Study:

• To develop and evaluate machine learning models for predicting COVID-19 outbreaks.
• To forecast the number of new COVID-19 cases over a 5-day period in highly populated countries.

Main Methods:

• Utilized 9 distinct machine learning algorithms to build prediction models.
• Focused on the 10 most highly and densely populated countries globally.
• Validated model performance with an average accuracy of 87.9% ± 3.9%.

Main Results:

• Developed a COVID-19 prediction system with an average accuracy of 87.9% ± 3.9% for the selected countries.
• Achieved a peak accuracy of 99.93% for Ethiopia using the Auto-Regressive Moving Average (ARMA) model over a 5-day forecast.
• The models demonstrated effectiveness in predicting the rise of new COVID-19 cases.

Conclusions:

• The developed prediction models can assist stakeholders in proactive preparation for sudden surges in COVID-19 cases.
• Effective resource management can be ensured through advance preparation facilitated by these predictive tools.
• The study highlights the utility of machine learning in managing public health crises like the COVID-19 pandemic.