Table 2 The parameters settings of CNN, CNN-attention and transformer model for prediction air consumption.
Parameter | CNN | CNN-Attention | Transformer |
|---|---|---|---|
Convolutional kernel size | 3 | 1 ~ 8 | None |
No. of convolutional kernels | 32\64\128 | 20 ~ 28 | |
No. of convolutional layers | 3 | 3 | |
Attention head count | None | 1 | 2 |
Embedding dimensions | None | 54 | 48 |
Epochs | 200 | 200 | 200 |
Learning rate | 0.01 | 0.007 | 0.003 |
Mini batch size | 32 | 32 | 32 |
Dropout rate | 0.1 | 0.1 | 0.1 |
Optimizer | Adam | Adam | Adam |
