Over-smoothing Effect of Graph Convolutional Networks

Fang Sun

Over-smoothing Effect of Graph Convolutional Networks

Machine Learning 2022-02-02 v2

Authors: Fang Sun

Abstract

Over-smoothing is a severe problem which limits the depth of Graph Convolutional Networks. This article gives a comprehensive analysis of the mechanism behind Graph Convolutional Networks and the over-smoothing effect. The article proposes an upper bound for the occurrence of over-smoothing, which offers insight into the key factors behind over-smoothing. The results presented in this article successfully explain the feasibility of several algorithms that alleviate over-smoothing.

Keywords

graph theory graph neural network graph algorithms

Cite

@article{arxiv.2201.12830,
  title  = {Over-smoothing Effect of Graph Convolutional Networks},
  author = {Fang Sun},
  journal= {arXiv preprint arXiv:2201.12830},
  year   = {2022}
}

Comments

8 pages, 2 figures

Related papers

View all related →