Table 1 Comparison of classification methods and their accuracy on NSL-KDD.

From: Anomaly-based intrusion detection on benchmark datasets for network security: a comprehensive evaluation

Method

Accuracy (%)

Naive Bayes Classifier8

97.00

K-Means + RF10

92.77

K-Means15

82.29

RNN20

89.6

Deep Stacking Network23

86.8

LSTM25

87.8

K-means clustering26

82.19

CNN27

93.65

CNN-BiLSTM hybrid31

98.27

LSTM43

97.54

Dugat-LSTM45

95

Enhanced DNN

99.96

Enhanced RNN

99.13

Back to article page