Back to papers
Faster Convergence in Mini-batch Graph Neural Networks Training with Pseudo Full Neighborhood Compensation
Summary: Identifies two gradient-estimation error sources in mini-batch neighbor-sampled GNNs—missing contributions from unsampled target nodes and inaccuracies in sampled-node messages—and shows prior work largely ignores the former. Proposes PFNC: a history-based, partial-cache, sampler-agnostic compensation scheme that corrects both errors, provably better approximates full-batch gradients, and empirically speeds convergence and improves generalization.
(summarized by gpt-5-mini on Feb 09 2026)
- Paper ID
- 14047
- Venue
- VLDB
- Year
- 2025
- Pagerank
- 4.1945683e-05
- Overall Rank
- 10,735 | 25.32%
- DOI
-
10.14778/3749646.3749695
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Incoming Citations (Sorted by Pagerank)
Showing 1 of 1 citing papers.
Outgoing Citations (Sorted by Pagerank)
Showing 17 of 17 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank |
Cited Paper |
Year |
Venue |
Pagerank |
| 278 |
AliGraph: A Comprehensive Graph Neural Network Platform |
2019 |
VLDB |
0.00029230623 |
| 1,160 |
Sancus: Staleness-Aware Communication-Avoiding Full-Graph Decentralized Training in Large-Scale Graph Neural Networks |
2022 |
VLDB |
0.00013586221 |
| 2,422 |
DUCATI: A Dual-Cache Training System for Graph Neural Networks on Giant Graphs with the GPU |
2023 |
SIGMOD |
8.8499665e-05 |
| 3,087 |
Scalable and Efficient Full-Graph GNN Training for Large Graphs |
2023 |
SIGMOD |
7.5939896e-05 |
| 5,018 |
DGC: Training Dynamic Graphs with Spatio-Temporal Non-Uniformity using Graph Partitioning by Chunks |
2023 |
SIGMOD |
5.7567672e-05 |
| 5,136 |
NeutronOrch: Rethinking Sample-based GNN Training under CPU-GPU Heterogeneous Environments |
2024 |
VLDB |
5.6723526e-05 |
| 5,321 |
FreshGNN: Reducing Memory Access via Stable Historical Embeddings for Graph Neural Network Training |
2024 |
VLDB |
5.5710441e-05 |
| 5,345 |
NeutronStream: A Dynamic GNN Training Framework with Sliding Window for Graph Streams |
2024 |
VLDB |
5.5567697e-05 |
| 5,475 |
ETC: Efficient Training of Temporal Graph Neural Networks over Large-scale Dynamic Graphs |
2024 |
VLDB |
5.4869706e-05 |
| 5,710 |
DynaHB: A Communication-Avoiding Asynchronous Distributed Framework with Hybrid Batches for Dynamic GNN Training |
2024 |
VLDB |
5.3590055e-05 |
| 5,737 |
Comprehensive Evaluation of GNN Training Systems: A Data Management Perspective |
2024 |
VLDB |
5.3480667e-05 |
| 6,884 |
Lotan: Bridging the Gap between GNNs and Scalable Graph Analytics Engines |
2023 |
VLDB |
4.8955332e-05 |
| 7,014 |
SIMPLE: Efficient Temporal Graph Neural Network Training at Scale with Dynamic Data Placement |
2024 |
SIGMOD |
4.8616315e-05 |
| 7,091 |
HongTu: Scalable Full-Graph GNN Training on Multiple GPUs |
2023 |
SIGMOD |
4.8370645e-05 |
| 7,289 |
DAHA: Accelerating GNN Training with Data and Hardware Aware Execution Planning |
2024 |
VLDB |
4.7747168e-05 |
| 7,566 |
ADGNN: Towards Scalable GNN Training with Aggregation-Difference Aware Sampling |
2023 |
SIGMOD |
4.7089968e-05 |
| 8,363 |
Eliminating Data Processing Bottlenecks in GNN Training over Large Graphs via Two-level Feature Compression |
2024 |
VLDB |
4.5364942e-05 |
Semantically Similar Papers
| Overall Rank |
Paper |
Year |
Venue |
Pagerank |
| 2,400 |
ByteGNN: Efficient Graph Neural Network Training at Large Scale |
2022 |
VLDB |
8.8955105e-05 |
| 7,566 |
ADGNN: Towards Scalable GNN Training with Aggregation-Difference Aware Sampling |
2023 |
SIGMOD |
4.7089968e-05 |
| 8,363 |
Eliminating Data Processing Bottlenecks in GNN Training over Large Graphs via Two-level Feature Compression |
2024 |
VLDB |
4.5364942e-05 |
| 8,740 |
Historical Embedding-Guided Efficient Large-Scale Federated Graph Learning |
2024 |
SIGMOD |
4.456315e-05 |
| 6,485 |
EARLY: Efficient and Reliable Graph Neural Network for Dynamic Graphs |
2023 |
SIGMOD |
5.0453531e-05 |
| 5,321 |
FreshGNN: Reducing Memory Access via Stable Historical Embeddings for Graph Neural Network Training |
2024 |
VLDB |
5.5710441e-05 |
| 5,737 |
Comprehensive Evaluation of GNN Training Systems: A Data Management Perspective |
2024 |
VLDB |
5.3480667e-05 |
| 3,087 |
Scalable and Efficient Full-Graph GNN Training for Large Graphs |
2023 |
SIGMOD |
7.5939896e-05 |
| 6,942 |
Efficient Training of Graph Neural Networks on Large Graphs |
2024 |
VLDB |
4.8922884e-05 |
| 10,539 |
Graph Neural Network Training Systems: A Performance Comparison of Full-Graph and Mini-Batch. |
2025 |
VLDB |
4.1945683e-05 |