Semi-Supervised Classification with Graph Convolutional Networks

Thomas N. Kipf, Max Welling
9/9/2016
cs.LGstat.ML

Official code: tkipf/pygcn (5,398 stars, MIT, last commit 2020-09-20). 1 more implementation below.

Abstract

We present a scalable approach for semi-supervised learning on graph-structured data that is based on an efficient variant of convolutional neural networks which operate directly on graphs. We motivate the choice of our convolutional architecture via a localized first-order approximation of spectral graph convolutions. Our model scales linearly in the number of graph edges and learns hidden layer representations that encode both local graph structure and features of nodes. In a number of experiments on citation networks and on a knowledge graph dataset we demonstrate that our approach outperforms related methods by a significant margin.

View on arXivView PDF

Code Implementations(2)

tkipf/pygcnOfficial100%
MIT

Graph Convolutional Networks in PyTorch

5,3981,214PythonOct 26, 20176 years agoMIT
tkipf/gcnCommunity100%
MIT

Implementation of Graph Convolutional Networks in TensorFlow

7,4121,995PythonNov 11, 20163 years agoMIT

Similar Papers

From Cluster Assumption to Graph Convolution: Graph-based Semi-Supervised Learning Revisited

88% match1 repo142023

FastGCN: Fast Learning with Graph Convolutional Networks via Importance Sampling

87% match1 repo5292018

Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering

85% match1 repo1.4k2016

Efficiently Learning the Graph for Semi-supervised Learning

85% match2023

Deep Graph Infomax

85% match1 repo6682018

Graph Random Neural Network for Semi-Supervised Learning on Graphs

85% match1 repo2102020

Cite this paper

@article{kipf2016semisupervised,
  title  = {Semi-Supervised Classification with Graph Convolutional Networks},
  author = {Thomas N. Kipf and Max Welling},
  year   = {2016},
  eprint = {1609.02907},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1609.02907}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1609.02907.svg)](https://codeofpaper.com/paper/1609.02907)

Discussion