Network Intrusion Detection with Nonsymmetric Deep Autoencoding Feature Extraction

Table 4

Experimental results of 4 attacks in NSL-KDD dataset.

Attack type	Precision (%)		Recall (%)		F1 (%)
	SNDAE	ANDAE	SNDAE	ANDAE	SNDAE	ANDAE
DoS	100.00	99.23	94.58	99.39	97.22	99.31
Probe	100.00	97.55	94.67	95.58	97.26	96.55
R2L	100.00	93.45	3.82	90.91	7.36	92.16
U2R	100.00	92.78	2.70	55.07	5.26	68.81

The bold values are the larger of the evaluation metrics of the two models.