Graph Dilated Network with Rejection Mechanism

Graph Dilated Network with Rejection Mechanism
复制标题

具有拒绝机制的图扩张网络

DOI:
10.3390/app10072421
复制
发表时间:
2020-04
影响因子:
2.7
通讯作者:
Guo Gaoyang
Guo Gaoyang
中科院分区:
综合性期刊4区
文献类型:
--
作者:
Yan Bencheng;Wang Chaokun;Guo Gaoyang

文献摘要

参考文献

相似文献

最近,图神经网络(GNNs)在处理基于图的数据方面取得了巨大成功。GNN的基本思想是迭代地聚合来自邻居的信息,这是拉普拉斯平滑的一种特殊形式。然而,大多数GNN都陷入了过度平滑问题,即,当模型深入时,学习的表示变得难以区分。这反映了当前GNN无法探索全局图结构。在本文中,我们提出了一种新的图神经网络来解决这个问题。设计了一种拒绝机制来解决过平滑问题,并提出了一种扩展的图卷积核来捕获高级图结构。大量的实验结果表明,该模型优于最先进的GNNs,并能有效地克服过平滑问题。
Recently, graph neural networks (GNNs) have achieved great success in dealing with graph-based data. The basic idea of GNNs is iteratively aggregating the information from neighbors, which is a special form of Laplacian smoothing. However, most of GNNs fall into the over-smoothing problem, i.e., when the model goes deeper, the learned representations become indistinguishable. This reflects the inability of the current GNNs to explore the global graph structure. In this paper, we propose a novel graph neural network to address this problem. A rejection mechanism is designed to address the over-smoothing problem, and a dilated graph convolution kernel is presented to capture the high-level graph structure. A number of experimental results demonstrate that the proposed model outperforms the state-of-the-art GNNs, and can effectively overcome the over-smoothing problem.
具有完全不平衡标签的网络嵌入
DOI: 10.1109/tkde.2020.2971490
发表时间: 2021
影响因子: 8.9
作者:
Zheng Wang;Ye Xiaojun;Wang Chaokun;Cui Jian;Yu Philip S.
通讯作者: Yu Philip S.
DOI: 10.1609/aaai.v32i1.11242
发表时间: 2018-04
期刊: --
影响因子: --
作者:
Zheng Wang;Xiaojun Ye;Chaokun Wang;Yuexin Wu;Changping Wang;Kaiwen Liang
通讯作者: Zheng Wang;Xiaojun Ye;Chaokun Wang;Yuexin Wu;Changping Wang;Kaiwen Liang
DOI: 10.1145/2939672.2939754
发表时间: 2016-08
期刊: KDD : proceedings. International Conference on Knowledge Discovery & Data Mining
影响因子: --
作者:
Grover A;Leskovec J
通讯作者: Leskovec J
DOI: --
发表时间: 2019-09
期刊: ArXiv
影响因子: --
作者:
Lingxiao Zhao;L. Akoglu
通讯作者: Lingxiao Zhao;L. Akoglu
NMLPA:通过多标签传播方法发现属性网络中的重叠社区
DOI: 10.3390/s19020260
发表时间: 2019-01
期刊: Sensors
影响因子: 3.9
作者:
Huang Bingyang;Wang Chaokun;Wang Binbin
通讯作者: Wang Binbin