Network Traffic Classification Based on SD Sampling and Hierarchical Ensemble Learning

Table 8

Model parameters of the two-layer structure.

Model	Parameters
	Original	SMOTE	Random sampling	SD sampling
XGBoost	n_estimators’: 90	“n_estimators”: 80	“n_estimators”: 80	“n_estimators”: 80
	max_depth’: 8	“max_depth”: 8	“max_depth”: 8	“max_depth”: 8
	learning rate’: 0.15	“learning_rate”: 0.2	“learning_rate”: 0.2	“learning_rate”: 0.2
	gamma’: 0.01	“gamma”: 0.001	“gamma”: 0.001	“gamma”: 0.001
RF	“n_estimators”: 20	“n_estimators”: 60	“n_estimators”: 60	“n_estimators”: 20
	“min_samples_leaf”: 1	“min_samples_leaf”: 1	“min_samples_leaf”: 1	“min_samples_leaf”: 1
	“max_features”: none	“max_features”: “log2”	“max_features”: “log2”	“max_features”: none