Threshold optimization and random undersampling for imbalanced credit card data

Journal of Big Data

Table 29 Results for the XGBoost depth 1 classifier

Technique	Threshold	TPR	FPR	FNR	TNR	F-meas	G-mean	MCC	Precision
F-meas	0.0678	0.9199	0.0399	0.0801	0.9601	0.0744	0.9396	0.1841	0.0388
F-meas NC	0.3362	0.8791	0.0054	0.1209	0.9946	0.3703	0.9349	0.4518	0.2390
G-mean	0.0607	0.9233	0.0459	0.0767	0.9541	0.0661	0.9384	0.1727	0.0343
G-mean NC	0.0883	0.9114	0.0308	0.0886	0.9692	0.0999	0.9396	0.2135	0.0531
MCC	0.0603	0.9239	0.0464	0.0761	0.9536	0.0657	0.9385	0.1721	0.0341
MCC NC	0.0894	0.9114	0.0305	0.0886	0.9695	0.1011	0.9398	0.2150	0.0539
Precision	0.0678	0.9199	0.0399	0.0801	0.9601	0.0744	0.9396	0.1841	0.0388
Precision NC	0.8462	0.7161	0.0007	0.2839	0.9993	0.6796	0.8386	0.6991	0.7259
C	0.1000	0.9083	0.0245	0.0917	0.9755	0.1139	0.9412	0.2312	0.0608
D	0.5000	0.8646	0.0026	0.1354	0.9974	0.5170	0.9284	0.5639	0.3708