GA-Stacking: A New Stacking-Based Ensemble Learning Method to Forecast the COVID-19 Outbreak

As a result of the increased number COVID-19 cases, Ensemble Machine Learning (EML) would be an effective tool for combatting this pandemic outbreak. An ensemble classifiers can improve performance single machine learning (ML) classifiers, especially stacking-based learning. Stacking utilizes heterogeneous-base learners trained in parallel and combines their predictions using meta-model to determine final prediction results. However, building often causes model decrease due increasing that are not being properly selected. Therefore, goal paper is develop evaluate generic, data-independent predictive method stacked-based (GA-Stacking) optimized by Genetic Algorithm (GA) outbreak health decision aided processes. GA-Stacking five well-known including Decision Tree (DT), Random Forest (RF), RIGID regression, Least Absolute Shrinkage Selection Operator (LASSO), eXtreme Gradient Boosting (XGBoost), at its first level. It also introduces GA identify comparisons forecast number, combination, trust these base based on Mean Squared Error (MSE) as fitness function. At second level stacked model, Linear Regression (LR) classifier used produce prediction. The was evaluated publicly available dataset from Center Systems Science Engineering, Johns Hopkins University, which consisted 10,722 data samples. experimental results indicated achieved outstanding with overall accuracy 99.99% three selected countries. Furthermore, proposed good when compared existing bagging-based approaches. predict correctly may applied generic epidemic trend other countries comparing preventive control measures.