Oversampling Imbalanced Data Based on Convergent WGAN for Network Threat Detection

Table 2

Details of the data sets.

Data sets	Component	Number of features	Number of samples (normal/abnormal)	IR
D1	Benign/malware	79	2709/1500	1.81
D2	Benign/defacement	79	2709/1300	2.08
D3	Benign/malware	79	2709/1000	2.71
D4	Benign/defacement	79	2709/800	3.39
D5	Benign/defacement	79	2709/500	5.42
D6	Benign/spam	79	2709/300	9.03
D7	Benign/spam	79	2709/140	19.35
D8	Benign/defacement	79	2709/54	50.17