Table 2. Confusion Matrix and Three Types of Errors of CatBoost, XGBoost, and LightGBM on the Test Set of Benchmark Data Set.

model name	TP	TN	FP	FN	ER	FPR	FNR
CatBoost	7339 (51.35%)	5516 (38.60%)	839 (5.87%)	598 (4.18%)	0.0587	0.1320	0.0753
XGBoost	7242 (50.67%)	5675 (39.71%)	680 (4.76%)	695 (4.86%)	0.0476	0.1070	0.0876
LightGBM	7302 (51.09%)	5513 (38.57%)	842 (5.89%)	635 (4.44%)	0.0589	0.1325	0.0800