Table 1 Example of model architecture and parameters
From: A novel adaptive momentum method for medical image classification using convolutional neural network
Layer | Output shape | Total parameter |
|---|---|---|
Convolution | 64 \(\times\) 64 \(\times\) 32 | 896 |
Batch Norm | 64 \(\times\) 64 \(\times\) 32 | 128 |
Dropout | 64 \(\times\) 64 \(\times 32\) | 0 |
Convolution | 62 \(\times\) 62 \(\times\) 64 | 18.496 |
Batch Norm | 62 \(\times\) 62 \(\times\) 64 | 256 |
Max Pool | 15 \(\times 15 \times 64\) | 0 |
Convolution | 15 \(\times 15 \times 128\) | 73,856 |
Batch Norm | 15 \(\times 15 \times 128\) | 512 |
Flatten | 28,800 | 0 |
Dense | 512 | 14.746.112 |
Batch | 512 | 2048 |
Dropout | 512 | 0 |
Dense | 4 | 516 |