Fig. 1
From: Learning compact graph representations via an encoder-decoder network
An example of the representation learning problem for graph classification. Here, the graph representation model is trained using the graphs attached to solid lines (G2 and G4). The corresponding representations (for G2 and G4) are then used to train the classifier. The trained model can then be used to generate representations for unseen samples, shown with dashed lines here (G1 and G3)