Threshold optimization and random undersampling for imbalanced credit card data

Journal of Big Data

Table 21 Results for the XGBoost depth 1 classifier

Technique	Threshold	TPR	FPR	FNR	TNR	F-meas	G-mean	MCC	Precision
F-meas	0.3385	0.9270	0.0623	0.0730	0.9377	0.0500	0.9322	0.1483	0.0257
F-meas NC	0.3647	0.9248	0.0570	0.0752	0.9430	0.0556	0.9337	0.1565	0.0287
G-mean	0.3201	0.9282	0.0666	0.0718	0.9334	0.0470	0.9307	0.1434	0.0241
G-mean NC	0.3670	0.9240	0.0565	0.0760	0.9435	0.0560	0.9335	0.1570	0.0289
MCC	0.3127	0.9291	0.0685	0.0709	0.9315	0.0458	0.9302	0.1414	0.0235
MCC NC	0.3810	0.9229	0.0540	0.0771	0.9460	0.0588	0.9343	0.1611	0.0304
Precision	0.3526	0.9248	0.0589	0.0752	0.9411	0.0525	0.9328	0.1522	0.0270
Precision NC	0.6989	0.8848	0.0176	0.1152	0.9824	0.1880	0.9318	0.2967	0.1215
C	0.5000	0.9120	0.0344	0.0880	0.9656	0.0854	0.9383	0.1975	0.0448
D	0.5000	0.9120	0.0344	0.0880	0.9656	0.0854	0.9383	0.1975	0.0448