Graph Classification with 2D Convolutional Neural Networks

TL;DR

This paper introduces a novel method to represent graphs as multi-channel images, enabling the use of standard 2D CNNs for graph classification, outperforming many existing methods in accuracy and efficiency.

Contribution

The paper proposes a simple yet effective graph-to-image transformation that allows vanilla 2D CNNs to be applied directly to graph classification tasks.

Findings

Outperforms state-of-the-art graph kernels and CNNs on 4 out of 6 datasets

Achieves comparable results on remaining datasets

Offers better time complexity than graph kernels

Abstract

Graph learning is currently dominated by graph kernels, which, while powerful, suffer some significant limitations. Convolutional Neural Networks (CNNs) offer a very appealing alternative, but processing graphs with CNNs is not trivial. To address this challenge, many sophisticated extensions of CNNs have recently been introduced. In this paper, we reverse the problem: rather than proposing yet another graph CNN model, we introduce a novel way to represent graphs as multi-channel image-like structures that allows them to be handled by vanilla 2D CNNs. Experiments reveal that our method is more accurate than state-of-the-art graph kernels and graph CNNs on 4 out of 6 real-world datasets (with and without continuous node attributes), and close elsewhere. Our approach is also preferable to graph kernels in terms of time complexity. Code and data are publicly available.