Table 1 ML-CNN composition and parameters
From: Fusion of transformer and ML-CNN-BiLSTM for network intrusion detection
Number | Network layer (module) | Size (depth) | Stride | Padding |
|---|---|---|---|---|
L1 | Convolutional layer | 14*14*1 | – | – |
M2 | M-Module | 64 | S = 1 | Same |
C3 | Convolutional layer | 4*4*32 | S = 1 | Same |
M4 | M-Module | 128 | S = 1 | Same |
C5 | Convolutional layer | 4*4*128 | S = 1 | Same |
M6 | M-Module | 256 | S = 1 | Same |
C7 | Average pooling layer | 2*2 | S = 1 | Same |
F8 | GAP layer | – | – | – |
R9 | Reshape layer | – | – | – |
D10 | LSTM layer | 128 | – | – |
F11 | Fully connected layer | 32 | – | – |