Mutual Information Maximization in Graph Neural Networks

Xinhan Di, Pengqian Yu, Rui Bu, Mingchao Sun
5/21/2019
cs.LGcs.CVstat.ML

Official code: CODE-SUBMIT/Graph_Neighborhood_1 (20 stars, last commit 2023-07-22). 1 more implementation below.

Abstract

A variety of graph neural networks (GNNs) frameworks for representation learning on graphs have been recently developed. These frameworks rely on aggregation and iteration scheme to learn the representation of nodes. However, information between nodes is inevitably lost in the scheme during learning. In order to reduce the loss, we extend the GNNs frameworks by exploring the aggregation and iteration scheme in the methodology of mutual information. We propose a new approach of enlarging the normal neighborhood in the aggregation of GNNs, which aims at maximizing mutual information. Based on a series of experiments conducted on several benchmark datasets, we show that the proposed approach improves the state-of-the-art performance for four types of graph tasks, including supervised and semi-supervised graph classification, graph link prediction and graph edge generation and classification.

View on arXivView PDF

Code Implementations(2)

CODE-SUBMIT/Graph_Neighborhood_1Official100%

First and Complementary Neighborhood Combination of Adjacency Matrix for Graph Learning

205PythonSep 3, 20193 years ago
CODE-SUBMIT/Neighborhood-Enlargement-in-Graph-NetworkOfficial100%

A Direct Neighborhood Combination of the Adjacency Matrix for Graph Learning

83PythonMay 20, 20197 years ago

Similar Papers

Simple and Efficient Heterogeneous Graph Neural Network

86% match1 repo1312022

Geom-GCN: Geometric Graph Convolutional Networks

86% match1 repo3112020

Global Self-Attention as a Replacement for Graph Convolution

85% match3 repos812021

Dynamic Graph Representation Learning via Self-Attention Networks

85% match1 repo3162018

Graph Coordinates and Conventional Neural Networks -- An Alternative for Graph Neural Networks

85% match2023

Neighbour-level Message Interaction Encoding for Improved Representation Learning on Graphs

84% match2024

Cite this paper

@article{di2019mutual,
  title  = {Mutual Information Maximization in Graph Neural Networks},
  author = {Xinhan Di and Pengqian Yu and Rui Bu and Mingchao Sun},
  year   = {2019},
  eprint = {1905.08509},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1905.08509}
}

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/1905.08509.svg)](https://codeofpaper.com/paper/1905.08509)

Discussion