Skip to main content

Table 3 Comparison of precision and recall for classification task on the Models

From: A DGA domain names detection modeling method based on integrating an attention mechanism and deep neural network

Domain TypeSVMCNNLSTMCNN-BiLSTMATT-CNN-BiLSTMSupport
 PRPRPRPRPR 
benign0.930.990.950.990.950.980.940.990.990.98199999
banjori0.950.970.990.980.99110.99115012
Cryptolocker000.110.120.250.240.220.020.280.281200
dyre0.920.980.990.990.9910.990.980.991167
emotet0.660.810.670.980.6710.6610.650.843985
gameover0.320.130.860.100.900.530.010.390.333323
locky0.270.050.560.050.6700.1400.460.281610
matsnu0.280.290.660.670.670.850.760.780.770.864133
murofet0.960.990.960.990.9410.970.990.9815304
Post0.610.720.790.730.760.750.680.870.850.764373
necurs0.250.240.540.220.520.340.570.190.610.684400
pykspa_v10.910.880.90.980.920.960.980.930.980.984662
qakbot0.630.430.740.610.710.640.620.730.760.675217
ramnit0.390.490.680.710.610.710.630.720.620.744900
rovnix0.850.910.99110.9910.9910.995160
suppobox0.750.320.850.230.860.220.990.070.860.942011
tinba0.580.810.690.890.680.960.740.980.920.983955
urlzone0.850.760.960.950.960.910.980.910.970.94369
volatile0.9710.980.710.970.6910.380.990.99185
beebone0.950.970.990.980.99110.991142
geodo0.850.760.960.950.960.910.980.910.980.93220
padcrypt0.640.420.740.610.710.640.620.730.760.66304
pizd0.800.790.800.820.890.890.990.980.960.96202
ramdo0.790.820.860.890.880.890.890.880.920.96400
shifu0.810.820.840.850.890.890.900.880.940.95510
micro avg0.780.790.830.830.840.840.860.860.890.89261643
macro avg0.640.640.650.640.690.650.720.720.830.83261643