Graph Neural Networks

Unifews: You Need Fewer Operations for Efficient Graph Neural Networks
Theoretical spectral bounds for GNN propagation and transformation sparsification.
Unifews: You Need Fewer Operations for Efficient Graph Neural Networks
SIGMA: An Efficient Heterophilous Graph Neural Network with Similarity-based Efficient Global Aggregation
Graph neural networks (GNNs) realize great success in graph learning but suffer from performance loss when meeting heterophily, i.e. …
Advances in Designing Scalable Graph Neural Networks: The Perspective of Graph Data Management
Graph Neural Network (GNN) is a successful marriage of graph data management and deep learning, leading to notable improvements in …
Advances in Designing Scalable Graph Neural Networks: The Perspective of Graph Data Management
Scalable Heterophilous Graph Neural Network with Decoupled Embeddings
We propose LD² as a scalable heterophilous GNN with linear complexity and lightweight minibatch training.
Jul 31, 2024 14:45 — 15:00 SMU, Singapore
Scalable Heterophilous Graph Neural Network with Decoupled Embeddings
GENTI: GPU-powered Walk-based Subgraph Extraction for Scalable Representation Learning on Dynamic Graphs
A GPU-oriented subgraph representation learning algorithm on dynamic data.
GENTI: GPU-powered Walk-based Subgraph Extraction for Scalable Representation Learning on Dynamic Graphs
Scalable Decoupling Graph Neural Networks with Feature-Oriented Optimization
SCARA extension with a more scalable reuse scheme.
Scalable Decoupling Graph Neural Networks with Feature-Oriented Optimization
LD²: Scalable Heterophilous Graph Neural Network with Decoupled Embeddings
A scalable heterophilous GNN with lightweight minibatch training.
LD²: Scalable Heterophilous Graph Neural Network with Decoupled Embeddings
Scaling up Graph Neural Networks
Qualifying Examinations report.
Scaling up Graph Neural Networks
SCARA: Scalable Graph Neural Networks with Feature-Oriented Optimization
A feature-oriented GNN that scales to billion-scale extremely large graphs.
SCARA: Scalable Graph Neural Networks with Feature-Oriented Optimization
Network Compression in Federated Machine Learning
Qualifying Examinations report.
Network Compression in Federated Machine Learning