Longer Time Span Air Pollution Prediction: The Attention and Autoencoder Hybrid Learning Model

Figure 12

Attention result on Shanghai dataset with node size of 64.