Table 7 The data output shape of each layer in CircCNN
From: Prediction of Back-splicing sites for CircRNA formation based on convolutional neural networks
SA Input | SD Input | |||
|---|---|---|---|---|
Type | Layer | Output Shape | Layer | Output Shape |
Input Layer | input_1 | (None, 100,4) | input_2 | (None, 100,4) |
Conv1D | conv1 | (None, 89,256) | conv3 | (None, 89,256) |
Dropout | dropout_1 | (None, 89,256) | dropout_4 | (None, 89,256) |
Conv1D | Conv2 | (None, 45,128) | conv4 | (None, 45,128) |
Dropout | dropout_2 | (None, 45,128) | dropout_5 | (None, 45,128) |
MaxPooling 1D | max_pooling 1d_1 | (None, 9128) | max_pooling 1d_2 | (None, 9128) |
Dropout | dropout_3 | (None, 9128) | dropout_6 | (None, 9128) |
Flatten | flatten_1 | (None, 1152) | flatten_2 | (None, 1152) |
Concatenate | cvout | (None, 2304) | ||
Batch Normalization | batch normalization_1 | (None, 2304) | ||