Threshold optimization and random undersampling for imbalanced credit card data

Journal of Big Data

Table 14 Results for the extremely randomized trees depth 8 classifier

Technique	Threshold	TPR	FPR	FNR	TNR	F-meas	G-mean	MCC	Precision
F-meas	0.0012	0.9087	0.0678	0.0913	0.9322	0.0446	0.9202	0.1380	0.0229
F-meas NC	0.1975	0.7735	0.0002	0.2265	0.9998	0.8143	0.8790	0.8159	0.8628
G-mean	0.0012	0.9114	0.0734	0.0886	0.9266	0.0414	0.9188	0.1326	0.0212
G-mean NC	0.0026	0.8807	0.0262	0.1193	0.9738	0.1164	0.9258	0.2274	0.0631
MCC	0.0011	0.9114	0.0737	0.0886	0.9263	0.0413	0.9186	0.1323	0.0211
MCC NC	0.0037	0.8725	0.0198	0.1275	0.9802	0.1639	0.9245	0.2715	0.0942
Precision	0.0012	0.9087	0.0678	0.0913	0.9322	0.0446	0.9202	0.1380	0.0229
Precision NC	0.6220	0.4836	0.0000	0.5164	1.0000	0.6437	0.6939	0.6844	0.9740
C	0.0017	0.8937	0.0390	0.1063	0.9610	0.0739	0.9266	0.1809	0.0386
D	0.5000	0.5652	0.0001	0.4348	0.9999	0.7047	0.7513	0.7275	0.9386