arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1911.09554cs.LGstat.ML
收藏

图上的离散和连续深度残差学习

Discrete and Continuous Deep Residual Learning Over Graphs

Universidade Federal do Rio Grande do Sul(南里奥格兰德联邦大学);Università Degli Studi di Siena(锡耶纳大学)

Pedro H. C. Avelar, Anderson R. Tavares, Marco Gori, Luis C. Lamb

摘要:

在本文中,我们提出了使用连续的剩余模块的图形内核的图神经网络。我们展示了离散和连续残差层如何允许更强大的训练,因为连续残差层是通过常微分方程(ODE)求解器集成以产生输出的残差层。我们的实验表明,当使用多层时,这些残差比使用非残差模块的残差获得更好的结果,减轻了基于GCN的模型的低通滤波效果。最后,我们应用和分析这些技术的行为,并指出这种技术如何在其他领域是有用的,允许更多的可预测的行为下的动态时间的计算。

英文摘要:

In this paper we propose the use of continuous residual modules for graph kernels in Graph Neural Networks. We show how both discrete and continuous residual layers allow for more robust training, being that continuous residual layers are those which are applied by integrating through an Ordinary Differential Equation (ODE) solver to produce their output. We experimentally show that these residuals achieve better results than the ones with non-residual modules when multiple layers are used, mitigating the low-pass filtering effect of GCN-based models. Finally, we apply and analyse the behaviour of these techniques and give pointers to how this technique can be useful in other domains by allowing more predictable behaviour under dynamic times of computation.

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。

↑