Timezone: »

 
Surfing on the Neural Sheaf
Julian Suk · Lorenzo Giusti · Tamir Hemo · Miguel Lopez · Marco La Vecchia · Konstantinos Barmpas · Cristian Bodnar

@
in Symmetry and Geometry in Neural Representations (NeurReps) »
Event URL: https://openreview.net/forum?id=xOXFkyRzTlu »

The deep connections between Partial Differential Equations (PDEs) and Graph Neural Networks (GNNs) have recently generated a lot of interest in PDE-inspired architectures for learning on graphs. However, despite being more interpretable and better understood via well-established tools from PDE analysis, the dynamics these models use are often too simple for complicated node classification tasks. The recently proposed Neural Sheaf Diffusion (NSD) models address this by making use of an additional geometric structure over the graph, called a sheaf, that can support a provably powerful class of diffusion equations. In this work, we propose Neural Sheaf Propagation (NSP), a new PDE-based Sheaf Neural Network induced by the wave equation on sheaves. Unlike diffusion models that are characterised by a dissipation of energy, wave models conserve energy, which can be beneficial for node classification tasks on heterophilic graphs. In practice, we show that NSP obtains competitive results with NSD and outperforms many other existent models.

Keywords: PDEs · Sheaf Neural Networks · wave equation

Author Information

Julian Suk (University of Twente)
Lorenzo Giusti (University of Roma "La Sapienza")
Tamir Hemo (California Institute of Technology)
Miguel Lopez (University of Pennsylvania)
Marco La Vecchia
Konstantinos Barmpas (Imperial College London)
Cristian Bodnar (University of Cambridge)

More from the Same Authors