Threshold optimization and random undersampling for imbalanced credit card data

Journal of Big Data

Table 16 Results for the random forest depth 4 classifier

Technique	Threshold	TPR	FPR	FNR	TNR	F-meas	G-mean	MCC	Precision
F-meas	0.0005	0.8929	0.0752	0.1071	0.9248	0.0400	0.9085	0.1285	0.0205
F-meas NC	0.3299	0.7547	0.0002	0.2453	0.9998	0.7997	0.8682	0.8019	0.8550
G-mean	0.0005	0.8996	0.0836	0.1004	0.9164	0.0362	0.9078	0.1220	0.0185
G-mean NC	0.0014	0.8612	0.0247	0.1388	0.9753	0.1632	0.9161	0.2644	0.0951
MCC	0.0005	0.8996	0.0836	0.1004	0.9164	0.0362	0.9078	0.1220	0.0185
MCC NC	0.0033	0.8524	0.0123	0.1476	0.9877	0.2766	0.9173	0.3654	0.1751
Precision	0.0005	0.8929	0.0752	0.1071	0.9248	0.0400	0.9085	0.1285	0.0205
Precision NC	0.8217	0.4545	0.0000	0.5455	1.0000	0.6190	0.6729	0.6659	0.9805
C	0.0017	0.8498	0.0089	0.1502	0.9911	0.2459	0.9176	0.3471	0.1440
D	0.5000	0.6898	0.0001	0.3102	0.9999	0.7832	0.8302	0.7906	0.9074