Breaking the Limits of Message Passing Graph Neural Networks

  Muhammet Balcilar 1 2 Pierre Heroux 1 Benoit Gauzere 3 Pascal Vasseur 1 4 Sebastien Adam 1 Paul Honeine 1

               Abstract                able weights. These weights can be shared with respect
     Since the Message Passing (Graph) Neural Net-       to the distance between nodes (Chebnet GNN) (Defferrard
     works (MPNNs) have a linear complexity with        et al., 2016), to the connected nodes features (GAT for graph
...