Research on DUAL-ADGAN Model for Anomaly Detection Method in Time-Series Data

Table 3

Ablation experiments results.

Dataset	Model	Acc	Pre	Recall	F1
RealAdExchange-CPC	Wad-GAN	0.941	0.857	0.75	0.8
	WganG_FenceD	0.948	0.8	0.923	0.857
	DUAL-ADGAN	0.962	0.848	0.967	0.903
RealTraffic-SPEED	Wad-GAN	0.912	0.734	0.647	0.688
	WganG_FenceD	0.921	0.786	0.647	0.71
	DUAL-ADGAN	0.949	0.818	0.9	0.857
RealTraffic-TravelTime	Wad-GAN	0.947	0.823	0.8	0.812
	WganG_FenceD	0.956	0.846	0.868	0.857
	DUAL-ADGAN	0.96	0.845	0.926	0.883