Graph neural networks (GNNs) have emerged recently as a powerful architecture for learning node and graph representations. Standard GNNs have the same expressive power as the Weisfeiler-Lehman test of graph isomorphism in terms of distinguishing non-isomorphic graphs. However, it was recently shown ...
4.2 K -hop peripheral-subgraph-enhanced graph neural network K-hop Peripheral-subgraph-enhanced Graph Neural Network (KP- GNN),核心部分在于 message passing 的过程中创造性的引入了其他 feature,并且更加 impressive,和传统的加残差这样 naive 的想法不同 \hat{h}_{v}^{l,k} = \mathbf{MES}_k^l(\...
论文信息 论文标题:How Powerful are K-hop Message Passing Graph Neural Networks论文作者:Jiarui Feng, Yixin Chen, Fuhai Li, Anindya Sarkar, Muhan Zhang论文来源:NeurIPS
具体可以参见NeurIPS 2021的论文“Decoupling the depth and scope of graph neural networks”arxiv 版本...