Order Graph Neural Network

Higher-order graph neural networks (HOGNNs) aim to improve upon the limitations of standard graph neural networks by incorporating information about higher-order relationships between nodes, going beyond simple pairwise connections. Current research focuses on enhancing the expressiveness and efficiency of HOGNNs, exploring architectures that leverage concepts like Clifford algebras, path complexes, and simplicial complexes to capture richer topological features and address issues like over-smoothing in high-degree graphs. These advancements are significant because they enable more accurate and robust predictions on complex graph-structured data, with applications spanning diverse fields including chemistry, physics, and social network analysis.

Papers

November 7, 2024

Higher-Order GNNs Meet Efficiency: Sparse Sobolev Graph Neural Networks
Jhony H. Giraldo, Aref Einizade, Andjela Todorovic, Jhon A. Castro-Correa, Mohsen Badiey, Thierry Bouwmans, Fragkiskos D. Malliaros
Graph Neural Network Graph Signal Convolution Operator Order Graph Neural Network Order Neighbor

October 7, 2024

A Clifford Algebraic Approach to E(n)-Equivariant High-order Graph Neural Networks
Hoang-Viet Tran, Thieu N. Vo, Tho Tran Huu, Tan Minh Nguyen
Inter Part Equivariance Geometric Deep Learning Equivariant Graph Neural Network Clifford Algebra Order Graph Neural Network

October 2, 2024

Rethinking the Expressiveness of GNNs: A Computational Model Perspective
Guanyu Cui, Zhewei Wei, Hsin-Hao Su
Graph Neural Network Graph Machine Learning Expressive Speech Message Passing Graph Neural Network Expressive GNN Order Graph Neural Network

June 18, 2024

Demystifying Higher-Order Graph Neural Networks
Maciej Besta, Florian Scheidl, Lukas Gianinazzi, Shachar Klaiman, Jürgen Müller, Torsten Hoefler
GNN Model GNN Architecture GNN Prediction Order Graph Order Graph Neural Network

June 12, 2024

Expressivity and Generalization: Fragment-Biases for Molecular GNNs
Tom Wollschläger, Niklas Kemper, Leon Hetzel, Johanna Sommer, Stephan Günnemann
Strong Generalization GNN Architecture Behavior Expressivity Style Order Graph Neural Network Molecular Graph Neural Network

June 8, 2024

Efficient Topology-aware Data Augmentation for High-Degree Graph Neural Networks
Yurui Lai, Xiaoyang Lin, Renchi Yang, Hongtao Wang
Graph Structured Data Graph Structure Graph Sparsification Order Graph Neural Network

November 28, 2023

PyTorch Geometric High Order: A Unified Library for High Order Graph Neural Network
Xiyuan Wang, Muhan Zhang
Pytorch Model Unified Library Subgraph GNN Order Graph Neural Network

August 13, 2023

Weisfeiler and Lehman Go Paths: Learning Topological Features via Path Complexes
Quang Truong, Peter Chin
Graph Neural Network Message Passing Topological Learning New Pathway Weisfeiler Leman Order Graph Neural Network

July 11, 2023

Influential Simplices Mining via Simplicial Convolutional Network
Yujie Zeng, Yiming Huang, Qiang Wu, Linyuan Lü
Higher Order Higher Order Network Influential Node Order Graph Neural Network Simplicial Convolutional Neural Network

June 29, 2023

Diffusion-Jump GNNs: Homophiliation via Learnable Metric Filters
Ahmed Begga, Francisco Escolano, Miguel Angel Lozano, Edwin R. Hancock
Metric Learning Diffusion Distance Order Graph Neural Network Smooth Latent

April 16, 2022

Theory of Graph Neural Networks: Representation and Learning
Stefanie Jegelka
Neural Network LeArning Abstract Graph Neural Network Theoretical Understanding Individual Representation Learning Model Order Graph Neural Network