mlsanityData to 7 Oct 2026

Model-Agnostic Augmentation for Accurate Graph Classification

arXiv 2202.10107 ↗First posted 21 Feb 2022cs.LG cs.SIcited by 21

Not on the map. Why?

Abstract (arXiv)

Given a graph dataset, how can we augment it for accurate graph classification? Graph augmentation is an essential strategy to improve the performance of graph-based tasks, and has been widely utilized for analyzing web and social graphs. However, previous works for graph augmentation either a) involve the target model in the process of augmentation, losing the generalizability to other tasks, or b) rely on simple heuristics that lead to unreliable results. In this work, we introduce five desired properties for effective augmentation. Then, we propose NodeSam (Node Split and Merge) and SubMix (Subgraph Mix), two model-agnostic approaches for graph augmentation that satisfy all desired properties with different motivations. NodeSam makes a balanced change of the graph structure to minimize the risk of semantic change, while SubMix mixes random subgraphs of multiple graphs to create rich soft labels combining the evidence for different classes. Our experiments on social networks and molecular graphs show that NodeSam and SubMix outperform existing approaches in graph classification.

Key facts. Cited by 21 papers in the map (data through 7 Oct 2026). Builds on 38 of its 58 references that are in the map. First posted on arXiv 21 Feb 2022.

Builds on

38 of 58 references.

References · 38

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,282
  3. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,155
  4. 2018How Powerful are Graph Neural Networks?cited by 4,355
  5. 2019CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Featurescited by 2,875
  6. 2017Improved Regularization of Convolutional Neural Networks with Cutoutcited by 1,623
  7. 2015DeepFool: a simple and accurate method to fool deep neural networkscited by 1,315
  8. 2017Random Erasing Data Augmentationcited by 1,235
  9. 2020Graph Contrastive Learning with Augmentationscited by 1,203
  10. 2018Graph Convolutional Neural Networks for Web-Scale Recommender Systemscited by 1,133
  11. 2018Hierarchical Graph Representation Learning with Differentiable Poolingcited by 786
  12. 2020TUDataset: A collection of benchmark datasets for learning with graphscited by 718
26 more
  1. 2019DropEdge: Towards Deep Graph Convolutional Networks on Node Classificationcited by 679
  2. 2019GraphSAINT: Graph Sampling Based Inductive Learning Methodcited by 618
  3. 2019Measuring and Relieving the Over-smoothing Problem for Graph Neural Networks from the Topological Viewcited by 544
  4. 2019Diffusion Improves Graph Learningcited by 414
  5. 2021A Survey of Data Augmentation Approaches for NLPcited by 408
  6. 2018Adversarial Attacks on Neural Networks for Graph Datacited by 395
  7. 2018Adversarial Attack on Graph Structured Datacited by 300
  8. 2021Graph Contrastive Learning Automatedcited by 287
  9. 2019Adversarial Attacks on Graph Neural Networks via Meta Learningcited by 261
  10. 2020Puzzle Mix: Exploiting Saliency and Local Statistics for Optimal Mixupcited by 260
  11. 2018GAN-based Synthetic Medical Image Augmentation for increased CNN Performance in Liver Lesion Classificationcited by 242
  12. 2020Time Series Data Augmentation for Deep Learning: A Surveycited by 225
  13. 2017Hide-and-Seek: Forcing a Network to be Meticulous for Weakly-supervised Object and Action Localizationcited by 207
  14. 2020Data Augmentation for Graph Neural Networkscited by 200
  15. 2019Topology Attack and Defense for Graph Neural Networks: An Optimization Perspectivecited by 196
  16. 2017Convolutional neural network architecture for geometric matchingcited by 138
  17. 2019Graph Convolutional Networks with EigenPoolingcited by 94
  18. 2019Graph Adversarial Training: Dynamically Regularizing Based on Graph Structurecited by 93
  19. 2018Data Augmentation using Random Image Cropping and Patching for Deep CNNscited by 87
  20. 2020Towards More Practical Adversarial Attacks on Graph Neural Networkscited by 76
  21. 2020GraphCrop: Subgraph Cropping for Graph Classificationcited by 31
  22. 2019GraphMix: Improved Training of GNNs for Semi-Supervised Learningcited by 23
  23. 2020Data Augmentation for Graph Classificationcited by 22
  24. 2016subgraph2vec: Learning Distributed Representations of Rooted Sub-graphs from Large Graphscited by 19
  25. 2017Data-Driven Color Augmentation Techniques for Deep Skin Image Analysiscited by 13
  26. 2020Signed Graph Diffusion Networkcited by 13

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 3; 2023: 7; 2024: 7; 2025: 3; 2026: 1.

21 papers.

Sort:

By area:

  1. 2023CoCo: A Coupled Contrastive Framework for Unsupervised Domain Adaptive Graph Classificationcited by 46
  2. 2022Unleashing the Power of Graph Data Augmentation on Covariate Distribution Shiftcited by 38
  3. 2023Data-centric Graph Learning: A Surveycited by 22
  4. 2023Graph Mixup with Soft Alignmentscited by 15
  5. 2023Fused Gromov-Wasserstein Graph Mixup for Graph-level Classificationscited by 15
  6. 2022Data Augmentation on Graphs: A Technical Surveycited by 10
  7. 2024GALA: Graph Diffusion-based Alignment with Jigsaw for Source-free Domain Adaptationcited by 8
  8. 2022Time-aware Random Walk Diffusion to Improve Dynamic Graph Learningcited by 7
  9. 2023Search to Capture Long-range Dependency with Stacking GNNs for Graph Classificationcited by 7
  10. 2024Accurate Link Prediction for Edge-Incomplete Graphs via PU Learningcited by 5
  11. 2024IntraMix: Intra-Class Mixup Generation for Accurate Labels and Neighborscited by 4
  12. 2023EPIC: Graph Augmentation with Edit Path Interpolation via Learnable Costcited by 3
9 more
  1. 2024Through the Dual-Prism: A Spectral Perspective on Graph Data Augmentation for Graph Classificationcited by 2
  2. 2024Graph Neural Network Generalization with Gaussian Mixture Model Based Augmentationcited by 2
  3. 2025AugWard: Augmentation-Aware Representation Learning for Accurate Graph Classificationcited by 2
  4. 2024TopER: Topological Embeddings in Graph Representation Learningcited by 1
  5. 2024Explanation-Preserving Augmentation for Semi-Supervised Graph Representation Learningcited by 1
  6. 2026TopoFormer: Topology Meets Attention for Graph Learningcited by 1
  7. 2023Model Stealing Attack against Graph Classification with Authenticity, Uncertainty and Diversitycited by 0
  8. 2025FracAug: Fractional Augmentation boost Graph-level Anomaly Detection under Limited Supervisioncited by 0
  9. 2025From Moments to Models: Graphon-Mixture Learning for Mixup and Contrastive Learningcited by 0