Table 9.

Comparison of different sampling methods.

Data set	Sampling method	Evaluation indicators (%)
		Accuracy	Precision	Recall	F1-score
NSL-KDD	Random oversampling	98.21	98.24	98.21	98.16
	Random undersampling	87.96	93.97	87.96	89.52
	SMOTE	98.77	98.85	98.77	98.81
	ADASYN	98.73	98.82	98.73	98.75
	ENN	97.62	98.13	97.62	97.87
	RENN	98.75	98.80	98.75	98.76
	ADRDB	99.81	99.76	99.81	99.79
UNSW_NB15	Random oversampling	73.87	84.94	73.87	77.08
	Random undersampling	61.12	73.66	61.12	63.00
	SMOTE	79.38	79.61	79.38	79.50
	ADASYN	80.74	80.14	80.74	80.14
	ENN	78.05	83.72	78.05	80.01
	RENN	78.69	83.31	78.69	79.22
	ADRDB	85.55	86.24	85.55	85.61
CIC-IDS2017	Random oversampling	92.13	91.87	92.13	89.82
	Random undersampling	34.72	90.95	34.72	39.76
	SMOTE	94.68	9.85	94.68	94.33
	ADASYN	95.58	94.88	95.58	95.02
	ENN	96.63	97.01	96.63	96.65
	RENN	97.31	97.51	97.31	97.41
	ADRDB	99.70	99.68	99.70	99.69