DEVELOPMENT OF ENSEMBLE MACHINE LEARNING MODEL TO IMPROVE COVID-19 OUTBREAK FORECASTING

Abstract

The world is currently facing the coronavirus disease 2019 (COVID-19) pandemic. Forecasting the progression of that pandemic is integral to planning the necessary next steps by governments and organizations. Recent studies have examined the factors that may impact COVID-19 forecasting, and others have built models for predicting the numbers of active cases, recovered cases, and deaths. The aim of this study was to improve the forecasting predictions by developing an ensemble machine learning model that can be utilized in addition to the Naïve Bayes classifier, which is one of the simplest and fastest probabilistic classifiers. The first ensemble model combined gradient boosting and random forest classifiers, and the second combined support vector machine and random forest classifiers. The numbers of confirmed, recovered, and death cases will be predicted for a period of 10 days. The results will be compared to the findings of the previous study. The results showed that the ensemble algorithm that combined gradient boosting and random forest achieved the best performance, with 0.99% accuracy in all cases. [JJCIT 2022; 8(2.000): 159-169]

Keywords

• covid-19
• coronavirus disease
• coronavirus
• pandemic
• epidemic prediction
• future forecasting
• machine learning
• ensemble machine learning algorithms
• naive bayes
• support vector machine
• random forest
• gradient boosting