Feature Transportation Improves Graph Neural Networks
This addresses a specific problem in graph-structured data analysis for researchers and practitioners, with incremental improvements over existing methods.
The paper tackled the challenge of modeling feature transportation in graph neural networks (GNNs) by proposing ADR-GNN, an architecture inspired by Advection-Diffusion-Reaction systems, which improved or offered competitive performance on node classification and spatio-temporal datasets.
Graph neural networks (GNNs) have shown remarkable success in learning representations for graph-structured data. However, GNNs still face challenges in modeling complex phenomena that involve feature transportation. In this paper, we propose a novel GNN architecture inspired by Advection-Diffusion-Reaction systems, called ADR-GNN. Advection models feature transportation, while diffusion captures the local smoothing of features, and reaction represents the non-linear transformation between feature channels. We provide an analysis of the qualitative behavior of ADR-GNN, that shows the benefit of combining advection, diffusion, and reaction. To demonstrate its efficacy, we evaluate ADR-GNN on real-world node classification and spatio-temporal datasets, and show that it improves or offers competitive performance compared to state-of-the-art networks.