An Evolutionary Computation Based Feature Selection Method for Intrusion Detection

Table 5

Classification accuracies of thresholds of SaDE in training sets in 4 datasets.

Dataset	0.6		0.7		0.8		0.5
	Mean	Std	Mean	Std	Mean	Std	Mean	Std
DataNum1	99.68	0.18	99.68	0.19	99.66	0.21	99.68	0.13
T-Sig	-		-		-
DataNum2	99.70	0.01	99.71	0.02	99.67	0.01	99.71	0.01
T-Sig	+		-		+
DataNum3	99.68	0.01	99.68	0.01	99.66	0.02	99.68	0.01
T-Sig	-		-		+
DataNum4	99.82	0.01	99.82	0.01	99.81	0.02	99.82	0.01
T-Sig	-		-		+