Threshold optimization and random undersampling for imbalanced credit card data

Journal of Big Data

Table 37 Results for the XGBoost depth 1 classifier

Technique	Threshold	TPR	FPR	FNR	TNR	F-meas	G-mean	MCC	Precision
F-meas	0.0077	0.9150	0.0414	0.0850	0.9586	0.0712	0.9364	0.1793	0.0370
F-meas NC	0.4845	0.8307	0.0004	0.1693	0.9996	0.7988	0.9110	0.7996	0.7716
G-mean	0.0072	0.9158	0.0455	0.0842	0.9545	0.0656	0.9348	0.1715	0.0340
G-mean NC	0.0110	0.9087	0.0293	0.0913	0.9707	0.1058	0.9390	0.2197	0.0566
MCC	0.0071	0.9166	0.0461	0.0834	0.9539	0.0648	0.9349	0.1704	0.0336
MCC NC	0.0117	0.9075	0.0279	0.0925	0.9721	0.1128	0.9391	0.2270	0.0607
Precision	0.0077	0.9150	0.0414	0.0850	0.9586	0.0712	0.9364	0.1793	0.0370
Precision NC	0.9753	0.4784	0.0001	0.5216	0.9999	0.6186	0.6827	0.6589	0.9326
C	0.0122	0.9040	0.0233	0.0960	0.9767	0.1184	0.9395	0.2357	0.0634
D	0.5000	0.8303	0.0004	0.1697	0.9996	0.8029	0.9108	0.8033	0.7785