A Residual Learning-Based Network Intrusion Detection System

Table 6

Comparison with other detection methods on UNSW-NB15 testing set.

Model	Testing metrics
	TN	TP	Acc	RW
RLF-CNN	36758	36278	0.887	0.893
RLC-CCNN	36996	35429	0.879	0.881
SMOTE-RF [32]	36952	32286	0.841	0.826
Pelican [19]	36850	34928	0.872	0.859
S-ResNet [1]	36928	31427	0.830	0.842