mlsanityData to 7 Oct 2026

Spectral Invariant Learning for Dynamic Graphs under Distribution Shifts

arXiv 2403.05026 ↗First posted 8 Mar 2024cs.LG cs.AIcited by 18

Not on the map. Why?

Abstract (arXiv)

Dynamic graph neural networks (DyGNNs) currently struggle with handling distribution shifts that are inherent in dynamic graphs. Existing work on DyGNNs with out-of-distribution settings only focuses on the time domain, failing to handle cases involving distribution shifts in the spectral domain. In this paper, we discover that there exist cases with distribution shifts unobservable in the time domain while observable in the spectral domain, and propose to study distribution shifts on dynamic graphs in the spectral domain for the first time. However, this investigation poses two key challenges: i) it is non-trivial to capture different graph patterns that are driven by various frequency components entangled in the spectral domain; and ii) it remains unclear how to handle distribution shifts with the discovered spectral patterns. To address these challenges, we propose Spectral Invariant Learning for Dynamic Graphs under Distribution Shifts (SILD), which can handle distribution shifts on dynamic graphs by capturing and utilizing invariant and variant spectral patterns. Specifically, we first design a DyGNN with Fourier transform to obtain the ego-graph trajectory spectrums, allowing the mixed dynamic graph patterns to be transformed into separate frequency components. We then develop a disentangled spectrum mask to filter graph dynamics from various frequency components and discover the invariant and variant spectral patterns. Finally, we propose invariant spectral filtering, which encourages the model to rely on invariant patterns for generalization under distribution shifts. Experimental results on synthetic and real-world dynamic graph datasets demonstrate the superiority of our method for both node classification and link prediction tasks under distribution shifts.

Key facts. Cited by 18 papers in the map (data through 7 Oct 2026). Builds on 59 of its 102 references that are in the map. First posted on arXiv 8 Mar 2024.

Builds on

59 of 102 references.

References · 59

  1. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  2. 2017Graph Attention Networkscited by 8,854
  3. 2013Efficient Estimation of Word Representations in Vector Spacecited by 4,346
  4. 2014Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translationcited by 4,016
  5. 2019Fast Graph Representation Learning with PyTorch Geometriccited by 2,414
  6. 2016Convolutional Neural Networks on Graphs with Fast Localized Spectral Filteringcited by 2,085
  7. 2019Invariant Risk Minimizationcited by 1,832
  8. 2016Variational Graph Auto-Encoderscited by 1,373
  9. 2019Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalizationcited by 1,232
  10. 2022FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecastingcited by 1,200
  11. 2021Domain Generalization: A Surveycited by 760
  12. 2020Out-of-Distribution Generalization via Risk Extrapolation (REx)cited by 737
38 more
  1. 2019Heterogeneous Graph Attention Networkcited by 732
  2. 2020Temporal Graph Networks for Deep Learning on Dynamic Graphscited by 521
  3. 2020Inductive Representation Learning on Temporal Graphscited by 454
  4. 2019EvolveGCN: Evolving Graph Convolutional Networks for Dynamic Graphscited by 443
  5. 2021Towards Out-Of-Distribution Generalization: A Surveycited by 431
  6. 2018Deep Learning on Graphs: A Surveycited by 393
  7. 2021Spectral Temporal Graph Neural Network for Multivariate Time-series Forecastingcited by 298
  8. 2021A Fourier-based Framework for Domain Generalizationcited by 287
  9. 2016Structured Sequence Modeling with Graph Convolutional Recurrent Networkscited by 273
  10. 2022Self-Supervised Contrastive Pre-Training For Time Series via Time-Frequency Consistencycited by 246
  11. 2021Multiwavelet-based Operator Learning for Differential Equationscited by 229
  12. 2020Focal Frequency Loss for Image Reconstruction and Synthesiscited by 204
  13. 2022Improving Out-of-Distribution Robustness via Selective Augmentationcited by 202
  14. 2020Invariant Risk Minimization Gamescited by 188
  15. 2021Inductive Representation Learning in Temporal Networks via Causal Anonymous Walkscited by 188
  16. 2022Discovering Invariant Rationales for Graph Neural Networkscited by 186
  17. 2022How Powerful are Spectral Graph Neural Networkscited by 169
  18. 2022Handling Distribution Shifts on Graphs: An Invariance Perspectivecited by 163
  19. 2019Graph Wavelet Neural Networkcited by 131
  20. 2020Invariant Rationalizationcited by 131
  21. 2022ROLAND: Graph Learning Framework for Dynamic Graphscited by 131
  22. 2020Foundations and modelling of dynamic networks using Dynamic Graph Neural Networks: A surveycited by 123
  23. 2021AdaRNN: Adaptive Learning and Forecasting of Time Seriescited by 120
  24. 2023Specformer: Spectral Graph Neural Networks Meet Transformerscited by 102
  25. 2022Out-Of-Distribution Generalization on Graphs: A Surveycited by 97
  26. 2019Variational Graph Recurrent Neural Networkscited by 86
  27. 2021Shift-Robust GNNs: Overcoming the Limitations of Localized Graph Training Datacited by 86
  28. 2022Wild-Time: A Benchmark of in-the-Wild Distribution Shift over Timecited by 85
  29. 2021OOD-GNN: Out-of-Distribution Generalized Graph Neural Networkcited by 78
  30. 2021Generalizing Graph Neural Networks on Out-Of-Distribution Graphscited by 69
  31. 2021Discrete-time Temporal Network Embedding via Implicit Hierarchical Learning in Hyperbolic Spacecited by 67
  32. 2020TeMP: Temporal Message Passing for Temporal Knowledge Graph Completioncited by 61
  33. 2020Structural Temporal Graph Neural Networks for Anomaly Detection in Dynamic Graphscited by 55
  34. 2020From Fourier to Koopman: Spectral Methods for Long-term Time Series Predictioncited by 50
  35. 2021Hyperbolic Variational Graph Neural Network for Modeling Dynamic Graphscited by 49
  36. 2022WOODS: Benchmarks for Out-of-Distribution Generalization in Time Seriescited by 29
  37. 2019Hierarchical Temporal Convolutional Networks for Dynamic Recommender Systemscited by 28
  38. 2021Interpretable Stability Bounds for Spectral Graph Filterscited by 28

50 most cited of 59 All 59 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 6; 2025: 8; 2026: 4.

18 papers.

Sort:

By area:

  1. 2024A Comprehensive Survey of Dynamic Graph Neural Networks: Models, Frameworks, Benchmarks, Experiments and Challengescited by 29
  2. 2024RealTCD: Temporal Causal Discovery from Interventional Data with Large Language Modelcited by 13
  3. 2024Graph Representation Learning via Causal Diffusion for Out-of-Distribution Recommendationcited by 11
  4. 2026USBD: Universal Structural Basis Distillation for Source-Free Graph Domain Adaptationcited by 9
  5. 2025Invariant Link Selector for Spatial-Temporal Out-of-Distribution Problemcited by 5
  6. 2024Graph Memory Learning: Imitating Lifelong Remembering and Forgetting of Brain Networkscited by 4
  7. 2026Self-Evolving Agents as Dynamic Graph Transformation: A Survey and New Perspectivecited by 4
  8. 2024A Survey of Out-of-distribution Generalization for Graph Machine Learning from a Causal Viewcited by 3
  9. 2025Evolving Graph Learning for Out-of-Distribution Generalization in Non-stationary Environmentscited by 2
  10. 2025Simple yet Effective Node Property Prediction on Edge Streams under Distribution Shiftscited by 1
  11. 2025Federated Causal Inference in Healthcare: Methods, Challenges, and Applicationscited by 1
  12. 2025Test-time GNN Model Evaluation on Dynamic Graphscited by 1
6 more
  1. 2025Towards Single-Source Domain Generalized Object Detection via Causal Visual Promptscited by 1
  2. 2026Towards OOD Generalization in Dynamic Graphs via Causal Invariant Learningcited by 1
  3. 2024Causal-aware Graph Neural Architecture Search under Distribution Shiftscited by 0
  4. 2025Robust Anomaly Detection Under Normality Distribution Shift in Dynamic Graphscited by 0
  5. 2025Debugging Tabular Log as Dynamic Graphscited by 0
  6. 2026SPR: Toward a Graph Foundation Model for Transferable Graph Cognition via Spectral Patterns and Relational Geometrycited by 0