Abstract
Graph Neural Networks (GNNs) have achieved significant success in modeling graph-structured data. However, their performance degrades on heterophilous graphs, where connected nodes tend to have dissimilar features. In this paper, we introduce the Dual-Channel Graph Neural Network (DCGNN), a novel GNN architecture designed to handle both homophily and heterophily by processing information through dual channels. DCGNN employs a homophily channel to capture information from similar nodes and a heterophily channel to learn from dissimilar nodes. We describe the detailed algorithm and provide a Python implementation using PyTorch. Extensive experimental results on benchmark heterophilous graph datasets demonstrate that DCGNN significantly outperforms standard GNN models and existing state-of-the-art methods.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
"Dual-Channel Graph Neural Network", Technical Disclosure Commons, (March 10, 2025)
https://www.tdcommons.org/dpubs_series/7885