Skip to main content

Table 6 Comparison results for the 10% KDD Cup 99 dataset

From: Entropy clustering-based granular classifiers for network intrusion detection

ClassifierTesting samplesTP/FPDDoSProbeU2RR2LMean TP/FP ratio 
FNT [27]11,982TP98.7598.3999.799.09203.84 
FP0.621.390.220.75
ID3 [28]311,029TP99.5297.8549.2192.75756.66 
FP0.040.550.1410.03 
Naïve Bayes [29]77,287TP79.294.812.20.1209.01 
FP1.713.30.90.3
SVM [30]15,437TP96.3263.8134.4885.6440.00 
FP1.630.940.054.4
J48 [31]15,437TP96.875.212.20.1148.75 
FP10.20.10.5
K-means [32]311,029TP97.387.629.86.4103.86 
FP0.42.60.40.1
SOM [31]15,437TP96.474.313.30.1125.35 
FP0.80.30.10.4
GAU [32]311,029TP82.490.222.89.660.28 
FP0.911.30.50.1
OneR [32]49,596TP94.212.910.710.763.80 
FP6.80.120.1
Bayes DT combo [29]311,029TP87.976.2312.3330.644.68 
FP0.671.78.923.8
NBTree [31]15,437TP97.473.31.20.140.00 
FP1.21.10.10.5
Naïve Baye s[31]15,437TP79.294.812.20.116.90 
FP1.713.30.90.3
The proposed ECGC15,437TP99.9897.8510097.573055 
FP0.010.0500.89