Deep Autoencoders and Feedforward Networks Based on a New Regularization for Anomaly Detection

Table 3

Comparison of the proposed models’ results with existing approaches used on the UNSW-NB15 dataset.

Model	No. of features	Classifier	Accuracy	FPR	TPR	Precision	F-score
Proposed FFN	42	DNN	94.7	1.04	94.24	86.76	89.75
Proposed VAE	42	Deep VAE	93.3	0.93	95.21	87.9	90.2
Two-stage ensemble [34]	—	Two-stage meta	91.72	8.90	91.30	91.60	—
GALR-DT [11]	20	DT	81.42	6.39	—	—	—
NAWIR [24]	42	AODE	83.47	6.57	98.5	—	—
[25]	5	RF	81.61	4.40	81.6	—	79.5
Standard MLP [37]	42	Softmax	81.30	21.15	—	—	—
TSDL [37]	10	Softmax	89.13	0.74	63.27	—	—
		NB	82.07	18.56
		DT	85.56	15.78
		ANN	81.34	21.13
		LR	83.15	18.48
[15]	42	EM	78.47	23.79	—	—	—