[Retracted] Transfer Learning Auto-Encoder Neural Networks for Anomaly Detection of DDoS Generating IoT Devices

Table 2

Percentage decrease in model accuracy before and after transfer-learning (TL) across device and malware type.

IoT device	Mirai (% decrease)		Bashlite (% decrease)
	Pre-TL	Post-TL	Pre-TL	Post-TL
DAD	18.83	0.82	40.04	1.25
ECT	22.15	1.37	44.65	4.24
P737	2.43	0.82	32.35	0.70
P838	2.51	0.14	32.01	1.00
S1003	8.79	1.03	33.77	−0.14
S1002	3.97	0.37	35.22	6.08
PBM	2.09	0.71	3.61	7.99
SNH	—	—	17.02	−0.12
END	—	—	37.07	0.01
Average (%)	8.68	0.75	30.64	2.33