A Framework of Rebalancing Imbalanced Healthcare Data for Rare Events’ Classification: A Case of Look-Alike Sound-Alike Mix-Up Incident Detection

Table 5

Comparison of classifiers with the best performance.

	Base classifier	Oversampling (% increase over LR)	Undersampling (%increase over LR)	SMOTE , (% increase over LR)
Recall	0.521	0.732 (40.50%)	0.555 (6.53%)	0.757 (45.30%)
Precision	0.694	0.577 (−16.86%)	0.598 (−13.83%)	0.597 (−13.98%)
F-score	0.595	0.645 (8.40%)	0.575 (−3.36%)	0.665 (11.76%)
Overall classification accuracy	0.850	0.829 (−2.47%)	0.826 (−2.82%)	0.837 (−1.53%)