Relational Pooling for Graph Representations
Relational Pooling for Graph Representations
复制标题
DOI：
--
复制
发表时间：
2019-03
期刊：
影响因子：
--
通讯作者：
R. Murphy;Balasubramaniam Srinivasan;Vinayak A. Rao;Bruno Ribeiro
中科院分区：
文献类型：
--
作者：
R. Murphy;Balasubramaniam Srinivasan;Vinayak A. Rao;Bruno Ribeiro
这项工作概括了图神经网络（GNN），超越了基于Weisfeiler-Lehman（WL）算法，图拉普拉斯算子和扩散的那些。我们的方法，表示关系池（RP），从有限的部分交换的理论，提供了一个框架，最大的代表性权力的图形。RP可以与现有的图表示模型一起工作，并且有点违反直觉，可以使它们比原始的WL同构测试更强大。此外，RP允许像递归神经网络和卷积神经网络这样的架构用于理论上合理的图分类方法。我们证明了基于RP的图形表示在许多任务上比最先进的方法性能更好。
This work generalizes graph neural networks (GNNs) beyond those based on the Weisfeiler-Lehman (WL) algorithm, graph Laplacians, and diffusions. Our approach, denoted Relational Pooling (RP), draws from the theory of finite partial exchangeability to provide a framework with maximal representation power for graphs. RP can work with existing graph representation models and, somewhat counterintuitively, can make them even more powerful than the original WL isomorphism test. Additionally, RP allows architectures like Recurrent Neural Networks and Convolutional Neural Networks to be used in a theoretically sound approach for graph classification. We demonstrate improved performance of RP-based graph representations over state-of-the-art methods on a number of tasks.