Skip to main navigation Skip to search Skip to main content

iPool - Information-Based Pooling in Hierarchical Graph Neural Networks

  • Xing Gao
  • , Wenrui Dai*
  • , Chenglin Li
  • , Hongkai Xiong
  • , Pascal Frossard
  • *Corresponding author for this work
  • Shanghai Jiao Tong University
  • Swiss Federal Institute of Technology Lausanne

Research output: Contribution to journal › Article › peer-review

Abstract

With the advent of data science, the analysis of network or graph data has become a very timely research problem. A variety of recent works have been proposed to generalize neural networks to graphs, either from a spectral graph theory or a spatial perspective. The majority of these works, however, focus on adapting the convolution operator to graph representation. At the same time, the pooling operator also plays an important role in distilling multiscale and hierarchical representations, but it has been mostly overlooked so far. In this article, we propose a parameter-free pooling operator, called iPool, that permits to retain the most informative features in arbitrary graphs. With the argument that informative nodes dominantly characterize graph signals, we propose a criterion to evaluate the amount of information of each node given its neighbors and theoretically demonstrate its relationship to neighborhood conditional entropy. This new criterion determines how nodes are selected and coarsened graphs are constructed in the pooling layer. The resulting hierarchical structure yields an effective isomorphism-invariant representation of networked data on arbitrary topologies. The proposed strategy achieves superior or competitive performance in graph classification on a collection of public graph benchmark data sets and superpixel-induced image graph data sets.

Original languageEnglish
Pages (from-to)5032-5044
Number of pages13
JournalIEEE Transactions on Neural Networks and Learning Systems
Volume33
Issue number9
DOIs
StatePublished - 1 Sep 2022
Externally publishedYes

Keywords

  • Graph classification
  • graph neural networks (GNNs)
  • graph pooling
  • hierarchical representation

Fingerprint

Dive into the research topics of 'iPool - Information-Based Pooling in Hierarchical Graph Neural Networks'. Together they form a unique fingerprint.

Cite this