BriefGPT.xyz
Nov, 2019
快速深度图神经网络
Fast and Deep Graph Neural Networks
HTML
PDF
Claudio Gallicchio, Alessio Micheli
TL;DR
本文提出了一种有效构建深度图神经网络的方法。该方法利用动力系统中固定点的思想，通过递归神经网络实现，并采用深度架构组织递归单元。我们证明了即使在没有训练递归连接的情况下，小型的深度图神经网络的性能也能达到或超过当前最先进的图分类任务水平。
Abstract
We address the
efficiency
issue for the construction of a
deep graph neural network
(GNN). The approach exploits the idea of representing each input graph as a fixed point of a dynamical system (implemented throu
→