Graph Neural Networks: Architectures, Stability, and Transferability
Graph Neural Networks: Architectures, Stability, and Transferability
复制标题
DOI：
10.1109/jproc.2021.3055400
复制
发表时间：
2021-05-01
影响因子：
20.6
通讯作者：
Ribeiro, Alejandro
中科院分区：
文献类型：
--
作者：
Ruiz, Luana;Gama, Fernando;Ribeiro, Alejandro
图神经网络（GNN）是用于图上支持的信号的信息处理架构。它们在这里作为卷积神经网络（CNN）的推广，其中各个层包含图形卷积滤波器组，而不是经典卷积滤波器组。否则，GNN就像CNN一样工作。滤波器是由逐点非线性和堆叠在层。结果表明，GNN结构表现出等变性置换和稳定的图变形。这些属性有助于解释GNN的良好性能，这些性能可以通过经验观察到。它还表明，如果图收敛到一个极限对象，一个graphon，GNNs收敛到一个相应的极限对象，一个graphon神经网络。这种收敛证明了GNN在具有不同节点数量的网络中的可转移性。通过将GNNs应用于推荐系统、分散式协作控制和无线通信网络来说明概念。
Graph neural networks (GNNs) are information processing architectures for signals supported on graphs. They are presented here as generalizations of convolutional neural networks (CNNs) in which individual layers contain banks of graph convolutional filters instead of banks of classical convolutional filters. Otherwise, GNNs operate as CNNs. Filters are composed of pointwise nonlinearities and stacked in layers. It is shown that GNN architectures exhibit equivariance to permutation and stability to graph deformations. These properties help explain the good performance of GNNs that can be observed empirically. It is also shown that if graphs converge to a limit object, a graphon, GNNs converge to a corresponding limit object, a graphon neural network. This convergence justifies the transferability of GNNs across networks with different numbers of nodes. Concepts are illustrated by the application of GNNs to recommendation systems, decentralized collaborative control, and wireless communication networks.