Current browse context:
cs
Change to browse by:
References & Citations
Computer Science > Machine Learning
Title: Semi-Supervised Classification with Graph Convolutional Networks
(Submitted on 9 Sep 2016 (v1), last revised 22 Feb 2017 (this version, v4))
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.
Submission history
From: Thomas Kipf [view email][v1] Fri, 9 Sep 2016 19:48:41 GMT (65kb,D)
[v2] Mon, 24 Oct 2016 21:25:47 GMT (792kb,D)
[v3] Thu, 3 Nov 2016 18:37:23 GMT (836kb,D)
[v4] Wed, 22 Feb 2017 09:55:36 GMT (858kb,D)
Link back to: arXiv, form interface, contact.