Machine Learning-Based Model to Predict the Disease Severity and Outcome in COVID-19 Patients

Table 10

Performance comparison of classifiers using top 10 features using original and SMOTE data.

Classifier	Sampling technique	Accuracy	Sensitivity	Specificity	F-score
LR	Without SMOTE	0.862	0.5	0.909	0.45
	With SMOTE	0.849	0.867	0.831	0.855
RF	Without SMOTE	0.89	0.63	0.934	0.609
	With SMOTE	0.925	0.884	0.983	0.933
XGB	Without SMOTE	0.851	0.455	0.908	0.43
	With SMOTE	0.89	0.843	0.965	0.904