Table 6 Final parameters of the Convolutional Neural Network architecture
Layer | Description | Values |
|---|---|---|
Input | ||
1 | (\(M \times 1 \times 1\)) | \(M=64\), 128 or 256 |
2 | Conv1D | \(T_1=4\) and \(Q_1=16\) |
3 | BachNorm | − |
4 | ReLu | − |
5 | MaxPool1D | \(S_1=8\) |
6 | Conv1D | \(T_2=4\) and \(Q_2=8\) |
7 | BachNorm | − |
8 | ReLu | − |
9 | MaxPool1D | \(S_2=8\) |
10 | Conv1D | \(T_3=2\) and \(Q_3=2\) |
11 | BachNorm | − |
12 | ReLu | − |
13 | MaxPool1D | \(S_3=2\) |
14 | Conv1D | \(T_4=2\) and \(Q_4=2\) |
15 | BachNorm | − |
16 | ReLu | − |
17 | MaxPool1D | \(S_4=2\) |
18 | FC1 | \(P_1=256\) |
19 | Dropout | \(\alpha _1=0.6\) |
20 | FC2 | \(P_2=128\) |
21 | Dropout | \(\alpha _2=0.6\) |
22 | FC3 | \(P_3=64\) |
23 | Dropout | \(\alpha _3=0.6\) |
24 | FC4 | \(P_4=2\) |
25 | SoftMax | 2 classes |