On Addressing the Limitations of Graph Neural Networks

Sitao Luan

On Addressing the Limitations of Graph Neural Networks

Machine Learning 2023-07-04 v2

Authors: Sitao Luan

Abstract

This report gives a summary of two problems about graph convolutional networks (GCNs): over-smoothing and heterophily challenges, and outlines future directions to explore.

Keywords

graph neural networks graph convolutional networks over-smoothing heterophily limitations

Cite

@article{arxiv.2306.12640,
  title  = {On Addressing the Limitations of Graph Neural Networks},
  author = {Sitao Luan},
  journal= {arXiv preprint arXiv:2306.12640},
  year   = {2023}
}

Comments

Proposal report and a concise version of Sitao Luan's thesis. The weight initialization part is quite interesting but will not be included in Sitao's formal thesis, thus Sitao put this preprint report online. arXiv admin note: substantial text overlap with arXiv:2109.05641, arXiv:2210.07606

Related papers

View all related →