Network Traffic Classification Based on SD Sampling and Hierarchical Ensemble Learning

Table 11

Classification performance of the sampled dataset with the random sampling algorithm.

	Precision	Recall	F1 score	FNR	FPR
Random sampling layer1
Benign	0.9993	0.9981	0.9987	0.0007	0.0019
Abnormal	0.9981	0.9993	0.9987	0.0019	0.0007
Accuracy					0.9987
Macro avg	0.9987	0.9987	0.9987	0.0013	0.0013
Weighted avg	0.9987	0.9987	0.9987	0.0013	0.0013
Random sampling layer2
Benign	0.9993	0.9981	0.9987	0.0007	0.0019
DoS hulk	0.9850	0.9992	0.9921	0.0150	0.0008
DDoS	0.9992	0.9996	0.9994	0.0008	0.0004
PortScan	1.0000	0.9984	0.9992	0.0000	0.0016
DoS goldeneye	0.9980	0.9972	0.9976	0.0020	0.0028
FTP-patator	1.0000	1.0000	1.0000	0.0000	0.0000
DoS slowloris	0.9948	0.9941	0.9944	0.0052	0.0059
DoS slowhttptest	0.9962	0.9946	0.9954	0.0038	0.0054
SSH-patator	1.0000	0.9963	0.9981	0.0000	0.0037
Bot	0.9979	0.9979	0.9979	0.0021	0.0021
Infiltration	1.0000	0.7778	0.8750	0.0000	0.2222
Heartbleed	1.0000	1.0000	1.0000	0.0000	0.0000
Web attack	0.9944	0.9869	0.9906	0.0056	0.0131
Accuracy					0.9977
Macro avg	0.9973	0.9800	0.9876	0.0027	0.0200
Weighted avg	0.9978	0.9977	0.9977	0.0022	0.0023