Table 1.

Impact of imbalance mitigation techniques on model performance.

Model	Adjustment	F1-Score	Recall	Precision	ROC-AUC
CatBoost	Unadjusted	0.83	0.85	0.81	0.86
	SMOTE + Class Weight	0.89	0.91	0.87	0.92
GradientBoosting	Unadjusted	0.76	0.79	0.75	0.81
	SMOTE + Class Weight	0.81	0.85	0.80	0.85
LinearSVC	Unadjusted	0.78	0.80	0.77	0.82
	Class Weighting	0.82	0.83	0.81	0.85
Random Forest	Unadjusted	0.80	0.82	0.78	0.83
	Stratified Sampling	0.86	0.87	0.85	0.89
Bernoulli Naïve Bayes	Unadjusted	0.78	0.83	0.77	0.82
	Class Weighting	0.82	0.88	0.80	0.85
Logistic Regression	Unadjusted	0.78	0.75	0.82	0.86
Logistic Regression	Class Weighting	0.84	0.8	0.88	0.89