GraphHD: Efficient graph classification using hyperdimensional computing

TL;DR

GraphHD introduces a hyperdimensional computing approach for graph classification, achieving comparable accuracy to GNNs with significantly faster training and inference, suitable for resource-constrained environments.

Contribution

This paper presents the first application of hyperdimensional computing to graph classification, demonstrating efficiency and competitive accuracy in IoT settings.

Findings

GraphHD achieves comparable accuracy to GNNs.

Training and inference are significantly faster.

Suitable for resource-limited environments.

Abstract

Hyperdimensional Computing (HDC) developed by Kanerva is a computational model for machine learning inspired by neuroscience. HDC exploits characteristics of biological neural systems such as high-dimensionality, randomness and a holographic representation of information to achieve a good balance between accuracy, efficiency and robustness. HDC models have already been proven to be useful in different learning applications, especially in resource-limited settings such as the increasingly popular Internet of Things (IoT). One class of learning tasks that is missing from the current body of work on HDC is graph classification. Graphs are among the most important forms of information representation, yet, to this day, HDC algorithms have not been applied to the graph learning problem in a general sense. Moreover, graph learning in IoT and sensor networks, with limited compute capabilities, introduce challenges to the overall design methodology. In this paper, we present GraphHD$-$a baseline approach for graph classification with HDC. We evaluate GraphHD on real-world graph classification problems. Our results show that when compared to the state-of-the-art Graph Neural Networks (GNNs) the proposed model achieves comparable accuracy, while training and inference times are on average 14.6$\times$ and 2.0$\times$ faster, respectively.