基于深度卷积神经网络的睡眠分期模型
|
|
贾子钰,林友芳,张宏钧,王晶
|
Sleep stage classification model based ondeep convolutional neural network
|
|
Zi-yu JIA,You-fang LIN,Hong-jun ZHANG,Jing WANG
|
|
| 表 2 预训练模型的最佳结构及详细参数 |
| Tab.2 The best structure and corresponding parameters of pretrained model |
|
| 层名称 | 层类型 | 单元数 | 激活函数 | 大小 | 步长 | | Input1 | − | − | − | − | − | | Con11 | Convolution | 64 | Relu | 50 | 6 | | MaxP11 | MaxPooling | − | − | 8 | 8 | | D11 | Dropout (0.5) | − | − | − | − | | Con12 | Convolution | 128 | Relu | 8 | 1 | | Con13 | Convolution | 128 | Relu | 8 | 1 | | Con14 | Convolution | 128 | Relu | 8 | 1 | | MaxP12 | MaxPooling | − | − | 4 | 4 | | F1 | Flatten | − | − | − | − | | Con21 | Convolution | 64 | Relu | 400 | 50 | | MaxP21 | MaxPooling | − | − | 4 | 4 | | D21 | Dropout (0.5) | − | − | − | − | | Con22 | Convolution | 128 | Relu | 6 | 1 | | Con23 | Convolution | 128 | Relu | 6 | 1 | | Con24 | Convolution | 128 | Relu | 6 | 1 | | MaxP22 | MaxPooling | − | − | 2 | 2 | | F2 | Flatten | − | − | − | − | | D3 | Dropout (0.5) | − | − | − | − | | Dense1 | Dense | 5 | Softmax | − | − |
|
|
|