A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf
.
Filters
Exploiting Neighbor Effect: Conv-Agnostic GNNs Framework for Graphs with Heterophily
[article]
2022
arXiv
pre-print
Moreover, we propose a Conv-Agnostic GNNs framework (CAGNNs) to enhance the performance of GNNs on heterophily datasets by learning the neighbor effect for each node. ...
Due to the homophily assumption in graph convolution networks, a common consensus is that graph neural networks (GNNs) perform well on homophilic graphs but may fail on heterophilic graphs with many inter-class ...
Motivated by this metric, we propose a Conv-Agnostic GNNs Framework (CAGNNs) to improve traditional GNNs' performance on heterophilic graphs by learning the node-level neighbor effect. ...
arXiv:2203.11200v2
fatcat:g3i3qyabavhbtflp3xpbi4htry
Universal Deep GNNs: Rethinking Residual Connection in GNNs from a Path Decomposition Perspective for Preventing the Over-smoothing
[article]
2022
arXiv
pre-print
Extensive experiments demonstrate the effectiveness of our method, which achieves state-of-the-art results over non-smooth heterophily datasets by simply stacking standard GNNs. ...
Based on these findings, we present a Universal Deep GNNs (UDGNN) framework with cold-start adaptive residual connections (DRIVE) and feedforward modules. ...
As shown in Figure 4 , where the encoder and decoder can be implemented with a linear layer, and the UDGNN is a conv-agnostic framework and can easily enhance the performance of GNNs in various datasets ...
arXiv:2205.15127v1
fatcat:sqxvn3iyenb3pgxi7aa4gyplie
On the power of message passing for learning on graph-structured data
2022
Lastly, we introduce PyTorch Geometric, a deep learning library for implementing and working with graph-based neural network building blocks, built upon PyTorch. ...
learning on graph-structured data, known as Graph Neural Networks (GNNs). ...
As a result, allowing GNNs to effectively operate in heterophily graphs (which requires the aggregation of informative representations from distant nodes) weakens the effects of over-smoothing and vice ...
doi:10.17877/de290r-22906
fatcat:mhcarapjczf2dbse7tirkcadwy