arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2210.09809cs.LG
收藏

使用神经正切核分析图神经网络中的卷积、非线性与深度

Analysis of Convolutions, Non-linearity and Depth in Graph Neural Networks using Neural Tangent Kernel

Technical University of Munich(慕尼黑工业大学)

Mahalakshmi Sabanayagam, Pascal Esser, Debarghya Ghoshdastidar

AI总结:

本文利用图神经正切核在半监督节点分类下理论分析GNN架构,证明线性网络与ReLU相当、行归一化最优、深度导致过度平滑但跳跃连接可消除该问题,并在真实数据集上验证。

摘要:

图神经网络(GNNs)的基本原理是通过“图卷积”聚合相邻节点,并结合网络架构的适当选择(如深度和激活函数)来利用数据的结构信息。因此,理解每个设计选择对网络性能的影响至关重要。基于图拉普拉斯算子的卷积已成为主导选择,其中邻接矩阵的对称归一化是最广泛采用的一种。然而,一些实证研究表明,邻接矩阵的行归一化在节点分类中优于对称归一化。尽管GNNs被广泛使用,但目前尚无严格的理论研究来解释这些卷积的表示能力,从而说明这一现象。类似地,线性GNNs性能与非线性ReLU GNNs相当的实证观察也缺乏严格的理论支持。\n在这项工作中,我们在半监督节点分类设置下,使用图神经正切核(Graph Neural Tangent Kernel)从理论上分析了GNN架构不同方面的影响。在总体度校正随机块模型(Degree Corrected Stochastic Block Model)下,我们证明:(i)线性网络捕获类别信息的能力与ReLU网络一样好;(ii)行归一化比其他卷积更好地保留了底层类别结构;(iii)由于过度平滑,性能随网络深度增加而下降,但行归一化中类别信息的损失最慢;(iv)跳跃连接即使在无限深度下也能保留类别信息,从而消除过度平滑。最后,我们通过数值实验以及在Cora和Citeseer等真实数据集上验证了我们的理论发现。

英文摘要:

The fundamental principle of Graph Neural Networks (GNNs) is to exploit the structural information of the data by aggregating the neighboring nodes using a `graph convolution' in conjunction with a suitable choice for the network architecture, such as depth and activation functions. Therefore, understanding the influence of each of the design choice on the network performance is crucial. Convolutions based on graph Laplacian have emerged as the dominant choice with the symmetric normalization of the adjacency matrix as the most widely adopted one. However, some empirical studies show that row normalization of the adjacency matrix outperforms it in node classification. Despite the widespread use of GNNs, there is no rigorous theoretical study on the representation power of these convolutions, that could explain this behavior. Similarly, the empirical observation of the linear GNNs performance being on par with non-linear ReLU GNNs lacks rigorous theory. In this work, we theoretically analyze the influence of different aspects of the GNN architecture using the Graph Neural Tangent Kernel in a semi-supervised node classification setting. Under the population Degree Corrected Stochastic Block Model, we prove that: (i) linear networks capture the class information as good as ReLU networks; (ii) row normalization preserves the underlying class structure better than other convolutions; (iii) performance degrades with network depth due to over-smoothing, but the loss in class information is the slowest in row normalization; (iv) skip connections retain the class information even at infinite depth, thereby eliminating over-smoothing. We finally validate our theoretical findings numerically and on real datasets such as Cora and Citeseer.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑