Table 2. Performance metrics of different grided models.

Model	Instance	Accuracy	Precision	Recall	F1-score	Training (s)	Testing (s)
RNN	Fraud	0.79	0.70	0.07	0.13	40.17	0.45
	Normal	0.79	0.80	0.99	0.88
LR	Fraud	0.94	0.81	0.94	0.96	1.13	0.00
	Normal	0.94	0.98	0.94	0.96
LOF	Fraud	0.77	0.37	0.09	0.14	8.50	0.19
	Normal	0.77	0.79	0.96	0.87
IF	Fraud	0.28	0.23	0.98	0.37	41.67	0.05
	Normal	0.28	0.93	0.08	0.15
SVM	Fraud	0.98	0.92	0.97	0.94	77.83	0.08
	Normal	0.98	0.99	0.98	0.98
RF	Fraud	0.99	0.98	0.95	0.97	1,572.28	0.03
	Normal	0.99	0.99	0.99	0.99
XGBoost	Fraud	0.99	0.96	0.98	0.97	408.53	0.00
	Normal	0.99	0.99	0.99	0.99