Search Results for author:
Found 3 papers, 0 papers with code
Steering Graph Neural Networks with Pinning Control
In the semi-supervised setting where labeled data are largely limited, it remains to be a big challenge for message passing based graph neural networks (GNNs) to learn feature representations for the nodes with the same class label that is distributed discontinuously over the graph.
Building Shortcuts between Distant Nodes with Biaffine Mapping for Graph Convolutional Networks
Multiple recent studies show a paradox in graph convolutional networks (GCNs), that is, shallow architectures limit the capability of learning information from high-order neighbors, while deep architectures suffer from over-smoothing or over-squashing.
Revisiting the role of heterophily in graph representation learning: An edge classification perspective
Alternatively, it is possible to exploit the information about the presence of heterophilous neighbors for feature learning, so a hybrid message passing approach is devised to aggregate homophilious neighbors and diversify heterophilous neighbors based on edge classification.
