Skip to main content

Table 87 Modes of XGBoost tuned hyperparameter values for experiments with the data exfiltration dataset

From: IoT information theft prediction using ensemble feature selection

Parameter name

Value

n_estimators

50

min_child_weight

3

max_depth

None

learning_rate

0.30000

gamma

None

  1. “None” value indicates default value of hyperparameter is optimal; parameter values for classifier yielding best results in terms of AUC