Network Traffic Classification Based on SD Sampling and Hierarchical Ensemble Learning

Table 10

Classification performance of the sampled dataset with the SMOTE algorithm.

	Precision	Recall	F1 score	FNR	FPR
SMOTE layer1
Benign	0.9999	0.9977	0.9988	0.0001	0.0023
Other	0.9977	0.9999	0.9988	0.0023	0.0001
Accuracy					0.9988
Macro avg	0.9988	0.9988	0.9988	0.0012	0.0012
Weighted avg	0.9988	0.9988	0.9988	0.0012	0.0012
SMOTE layer2
Benign	0.9999	0.9977	0.9988	0.0001	0.0023
DoS hulk	0.9819	1.0000	0.9909	0.0181	0.0000
DDoS	1.0000	1.0000	1.0000	0.0000	0.0000
PortScan	1.0000	1.0000	1.0000	0.0000	0.0000
DoS goldeneye	0.9988	0.9980	0.9984	0.0012	0.0020
FTP-patator	1.0000	1.0000	1.0000	0.0000	0.0000
DoS slowloris	0.9955	0.9941	0.9948	0.0045	0.0059
DoS slowhttptest	0.9962	0.9946	0.9954	0.0038	0.0054
SSH-patator	1.0000	0.9975	0.9988	0.0000	0.0025
Bot	1.0000	0.9979	0.9990	0.0000	0.0021
Infiltration	1.0000	0.8889	0.9412	0.0000	0.1111
Heartbleed	1.0000	1.0000	1.0000	0.0000	0.0000
Web attack	0.9925	0.9907	0.9916	0.0075	0.0093
Accuracy					0.9979
Macro avg	0.9973	0.9892	0.9930	0.0027	0.0108
Weighted avg	0.9980	0.9979	0.9979	0.0020	0.0021