Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Beyond Generalization: A Survey of Out-Of-Distribution Adaptation on Graphs

Shuhan Liu, Kaize Ding

TL;DR

Addressing graph distribution shifts, the paper surveys graph OOD adaptation across training-time and test-time settings, formalizes problem definitions with source and target distributions, and develops a two-axis taxonomy (model-centric vs data-centric; training-time vs test-time). It reviews representative methods including domain-invariant representation learning, adversarial alignment, and graph transformations, as well as techniques for open-set and concept-shift, providing a synthesis of strengths and limitations. The work clarifies connections to graph transfer learning, domain adaptation, and fair learning, and offers guidance on when to apply which family of methods. It also outlines key challenges and future directions, such as theoretical foundations, scalability to complex/temporal graphs, and robust evaluation in real-world networks.

Abstract

Distribution shifts on graphs -- the data distribution discrepancies between training and testing a graph machine learning model, are often ubiquitous and unavoidable in real-world scenarios. Such shifts may severely deteriorate the performance of the model, posing significant challenges for reliable graph machine learning. Consequently, there has been a surge in research on graph Out-Of-Distribution (OOD) adaptation methods that aim to mitigate the distribution shifts and adapt the knowledge from one distribution to another. In our survey, we provide an up-to-date and forward-looking review of graph OOD adaptation methods, covering two main problem scenarios including training-time as well as test-time graph OOD adaptation. We start by formally formulating the two problems and then discuss different types of distribution shifts on graphs. Based on our proposed taxonomy for graph OOD adaptation, we systematically categorize the existing methods according to their learning paradigm and investigate the techniques behind them. Finally, we point out promising research directions and the corresponding challenges. We also provide a continuously updated reading list at https://github.com/kaize0409/Awesome-Graph-OOD-Adaptation.git

Beyond Generalization: A Survey of Out-Of-Distribution Adaptation on Graphs

TL;DR

Addressing graph distribution shifts, the paper surveys graph OOD adaptation across training-time and test-time settings, formalizes problem definitions with source and target distributions, and develops a two-axis taxonomy (model-centric vs data-centric; training-time vs test-time). It reviews representative methods including domain-invariant representation learning, adversarial alignment, and graph transformations, as well as techniques for open-set and concept-shift, providing a synthesis of strengths and limitations. The work clarifies connections to graph transfer learning, domain adaptation, and fair learning, and offers guidance on when to apply which family of methods. It also outlines key challenges and future directions, such as theoretical foundations, scalability to complex/temporal graphs, and robust evaluation in real-world networks.

Abstract

Distribution shifts on graphs -- the data distribution discrepancies between training and testing a graph machine learning model, are often ubiquitous and unavoidable in real-world scenarios. Such shifts may severely deteriorate the performance of the model, posing significant challenges for reliable graph machine learning. Consequently, there has been a surge in research on graph Out-Of-Distribution (OOD) adaptation methods that aim to mitigate the distribution shifts and adapt the knowledge from one distribution to another. In our survey, we provide an up-to-date and forward-looking review of graph OOD adaptation methods, covering two main problem scenarios including training-time as well as test-time graph OOD adaptation. We start by formally formulating the two problems and then discuss different types of distribution shifts on graphs. Based on our proposed taxonomy for graph OOD adaptation, we systematically categorize the existing methods according to their learning paradigm and investigate the techniques behind them. Finally, we point out promising research directions and the corresponding challenges. We also provide a continuously updated reading list at https://github.com/kaize0409/Awesome-Graph-OOD-Adaptation.git
Paper Structure (13 sections, 1 equation, 2 figures, 1 table)

This paper contains 13 sections, 1 equation, 2 figures, 1 table.

Table of Contents

  1. Introduction
  2. Graph Out-Of-Distribution Adaptation
  3. Problem Definition
  4. Discussion on Related Topics
  5. Taxonomy
  6. Training-Time Graph OOD Adaptation
  7. Model-Centric Approaches
  8. Data-Centric Approaches
  9. Test-Time Graph OOD Adaptation
  10. Model-Centric Approaches
  11. Data-Centric Approaches
  12. Future Directions
  13. Conclusion

Figures (2)

  • Figure 1: An illustration of training-time graph OOD adaptation and test-time graph OOD adaptation.
  • Figure 2: Overview of our proposed taxonomy