期刊
SCIENCE CHINA-INFORMATION SCIENCES
卷 66, 期 12, 页码 -出版社
SCIENCE PRESS
DOI: 10.1007/s11432-021-3443-y
关键词
graph neural networks; receptive field; reinforcement learning
In graph convolutional networks, the importance of different nodes in a neighborhood varies. We propose a framework of conducting graph convolutions using adaptive receptive fields (ARFs) to address the smoothness issue of soft attention weights and efficiently explore long-distance dependencies. We further introduce GRARF as an instance, which achieves state-of-the-art performances and shows more robustness against neighborhood noises compared to attention models.
Different nodes in a graph neighborhood generally yield different importance. In previous work of graph convolutional networks (GCNs), such differences are typically modeled with attention mechanisms. However, as we prove in our paper, soft attention weights suffer from undesired smoothness large neighborhoods (not to be confused with the oversmoothing effect in deep GCNs). To address this weakness, we introduce a novel framework of conducting graph convolutions, where nodes are discretely selected among multi-hop neighborhoods to construct adaptive receptive fields (ARFs). ARFs enable GCNs to get rid of the smoothness of soft attention weights, as well as to efficiently explore long-distance dependencies in graphs. We further propose GRARF (GCN with reinforced adaptive receptive fields) as an instance, where an optimal policy of constructing ARFs is learned with reinforcement learning. GRARF achieves or matches state-of-the-art performances on public datasets from different domains. Our further analysis corroborates that GRARF is more robust than attention models against neighborhood noises.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据