Node Sampling

Node sampling techniques aim to efficiently process large graphs by selecting representative subsets of nodes for graph neural network (GNN) training and inference. Current research focuses on developing adaptive sampling methods that go beyond static heuristics, incorporating techniques like Bayesian networks, transformers, and random walks to improve sampling efficiency and accuracy, particularly for heterophilous graphs and distributed settings. These advancements are crucial for scaling GNNs to massive datasets, enabling their application in diverse fields like recommendation systems and financial modeling where graph data is prevalent. The ultimate goal is to achieve high accuracy with minimal computational cost and communication overhead.

Papers

October 20, 2024

Efficient Model Extraction via Boundary Sampling
Maor Biton Dor, Yisroel Mirsky
Adversarial Example Black Box Black Box Model Extraction Attack Model Extraction Node Sampling

February 3, 2024

ParZC: Parametric Zero-Cost Proxies for Efficient NAS
Peijie Dong, Lujun Li, Xinglin Pan, Zimian Wei, Xiang Liu, Qiang Wang, Xiaowen Chu
Neural Architecture Search Zero Cost Proxy Zero Shot Neural Architecture Search Node Sampling

October 31, 2023

Diversified Node Sampling based Hierarchical Transformer Pooling for Graph Representation Learning
Gaichao Li, Jinsong Chen, John E. Hopcroft, Kun He
Graph Representation Learning Graph Pooling Graph Task Graph Pooling Method Node Sampling

October 5, 2023

GRAPES: Learning to Sample Graphs for Scalable Graph Neural Networks
Taraneh Younesian, Daniel Daza, Emile van Krieken, Thiviyan Thanapalasingam, Peter Bloem
Graph Neural Network Node Classification Heterophilic Graph Node Sampling Table Grape Graph Sample

May 4, 2023

Communication-Efficient Graph Neural Networks with Probabilistic Neighborhood Expansion Analysis and Caching
Tim Kaler, Alexandros-Stavros Iliopoulos, Philip Murzynowski, Tao B. Schardl, Charles E. Leiserson, Jie Chen
Graph Neural Network Forward Caching Open Graph Benchmark Node Sampling

November 16, 2022

Hierarchical Estimation for Effective and Efficient Sampling Graph Neural Network
Yang Li, Bingbing Xu, Qi Cao, Yige Yuan, Huawei Shen
Graph Neural Network Node Embeddings Hierarchical Bayesian Neighbor Sampling Subgraph Sampling Node Sampling

October 8, 2022

Hierarchical Graph Transformer with Adaptive Node Sampling
Zaixi Zhang, Qi Liu, Qingyong Hu, Chee-Kong Lee
Transformer Architecture Graph Transformer Node Sampling

June 1, 2022

Walk for Learning: A Random Walk Approach for Federated Learning from Heterogeneous Data
Ghadir Ayache, Venkat Dassari, Salim El Rouayheb
LeArning Abstract Heterogeneous Data Random Walk Optimal Sampling Seated Walk Adaptive Diffusion Node Sampling

March 21, 2022

BNS-GCN: Efficient Full-Graph Training of Graph Convolutional Networks with Partition-Parallelism and Random Boundary Node Sampling
Cheng Wan, Youjie Li, Ang Li, Nam Sung Kim, Yingyan Lin
Large Scale Graph C2p GCN Level Parallelism Node Sampling

Node Sampling

Papers

Efficient Model Extraction via Boundary Sampling

ParZC: Parametric Zero-Cost Proxies for Efficient NAS

Diversified Node Sampling based Hierarchical Transformer Pooling for Graph Representation Learning

GRAPES: Learning to Sample Graphs for Scalable Graph Neural Networks

Communication-Efficient Graph Neural Networks with Probabilistic Neighborhood Expansion Analysis and Caching

Hierarchical Estimation for Effective and Efficient Sampling Graph Neural Network

Hierarchical Graph Transformer with Adaptive Node Sampling

Walk for Learning: A Random Walk Approach for Federated Learning from Heterogeneous Data

BNS-GCN: Efficient Full-Graph Training of Graph Convolutional Networks with Partition-Parallelism and Random Boundary Node Sampling