Table 5 The hyper-parameters of the proposed CNN based on the default mode.
From: A deep learning framework combined with word embedding to identify DNA replication origins
Hyper-parameter | Value |
|---|---|
Input length | \(L_{max}\) |
Batch size | 64 |
Embedding layer | \(L_{max} \times 300\), Trainable = False, |
Num_Channels = 1 | |
Convolution blocks | \([2,\ 3,\ 4]\), \(128 \times 3\), ReLu |
Pooling blocks | Max-pooling |
Fully connected layer units | \(128 \times 3\) |
Regularization | L2 |
Learning rate | 0.001 with decay rate 0.9 |
Optimizer | Adam |
