Research Article
ECG Signal Classification Based on Fusion of Hybrid CNN and Wavelet Features by D-S Evidence Theory
Table 3
The architecture of 1-D CNN.
| Layers | Type | Activation function | Number of neurons (output layer) | Filter size | Number of filters | Stride |
| 1 | Input | | | | | | 2 | Convolution | ReLU | | | 4 | 1 | 3 | Max-pooling | | | | 4 | 2 | 4 | Convolution | ReLU | | | 4 | 1 | 5 | Max-pooling | | | | 4 | 2 | 6 | Convolution | ReLU | | | 4 | 1 | 7 | Max-pooling | | | | 4 | 2 | 8 | Convolution | ReLU | | | 8 | 1 | 9 | Max-pooling | | | | 8 | 2 | 10 | Fully connected | ReLU | 10 | | | | 11 | Output | SoftMax | 5 | | | |
|
|