Learning Convolutional Neural Networks for Graphs

TL;DR

This paper introduces a flexible framework for applying convolutional neural networks to various types of graph data, enabling efficient and competitive feature learning across different graph structures.

Contribution

It presents a novel method for defining convolutional operations on arbitrary graphs, extending CNN applicability beyond images.

Findings

Competitive performance with state-of-the-art graph kernels

High computational efficiency

Effective feature representation learning

Abstract

Numerous important problems can be framed as learning from graph data. We propose a framework for learning convolutional neural networks for arbitrary graphs. These graphs may be undirected, directed, and with both discrete and continuous node and edge attributes. Analogous to image-based convolutional networks that operate on locally connected regions of the input, we present a general approach to extracting locally connected regions from graphs. Using established benchmark data sets, we demonstrate that the learned feature representations are competitive with state of the art graph kernels and that their computation is highly efficient.