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

Figure 10

GRU result on Shanghai dataset with node size of 64.