Advances, Systems and Applications
Table 2 Adopted values for each parameters of PERCNN
From: Ground radar precipitation estimation with deep learning approaches in meteorological private cloud
Parameter | Values |
|---|---|
Convolutional layers | 2 |
Filter | 5×5 |
Stride | 1 |
Pooling layers | #Pool1 None |
| Â | #Pool2 None |
Learning rate | 0.0001 |
Learning decay rate | 0.2 |
Activation function | ReLU |
Optimizer | Adam |