Table 7. Classification performance on original NSL-KDD dataset (41 features).

Classifier	Accuracy	Precision	Recall	F1-measure
Random forest	71.91%	68.76%	72.66%	70.65%
KNN	68.71%	67.61%	69.25%	68.42%
SVM	74.32%	72.19%	69.92%	71.03%
Decision tree	67.39%	66.21%	65.43%	65.81%
MLP	78.32%	79.43%	77.14%	78.26%
Ensemble	82.91%	81.82%	84.78%	83.27%