Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
129 tokens/sec
GPT-4o
28 tokens/sec
Gemini 2.5 Pro Pro
42 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

GOODAT: Towards Test-time Graph Out-of-Distribution Detection (2401.06176v1)

Published 10 Jan 2024 in cs.LG and cs.AI

Abstract: Graph neural networks (GNNs) have found widespread application in modeling graph data across diverse domains. While GNNs excel in scenarios where the testing data shares the distribution of their training counterparts (in distribution, ID), they often exhibit incorrect predictions when confronted with samples from an unfamiliar distribution (out-of-distribution, OOD). To identify and reject OOD samples with GNNs, recent studies have explored graph OOD detection, often focusing on training a specific model or modifying the data on top of a well-trained GNN. Despite their effectiveness, these methods come with heavy training resources and costs, as they need to optimize the GNN-based models on training data. Moreover, their reliance on modifying the original GNNs and accessing training data further restricts their universality. To this end, this paper introduces a method to detect Graph Out-of-Distribution At Test-time (namely GOODAT), a data-centric, unsupervised, and plug-and-play solution that operates independently of training data and modifications of GNN architecture. With a lightweight graph masker, GOODAT can learn informative subgraphs from test samples, enabling the capture of distinct graph patterns between OOD and ID samples. To optimize the graph masker, we meticulously design three unsupervised objective functions based on the graph information bottleneck principle, motivating the masker to capture compact yet informative subgraphs for OOD detection. Comprehensive evaluations confirm that our GOODAT method outperforms state-of-the-art benchmarks across a variety of real-world datasets. The code is available at Github: https://github.com/Ee1s/GOODAT

Definition Search Book Streamline Icon: https://streamlinehq.com
References (60)
  1. Deep Variational Information Bottleneck. In ICLR 2017.
  2. Feed Two Birds with One Scone: Exploiting Wild Data for Both Out-of-Distribution Generalization and Detection. In ICML 2023.
  3. LOF: Identifying Density-Based Local Outliers. In SIGMOD 2000.
  4. A topological proof of Sklar’s theorem. Applied Mathematics Letters, 26(9): 945–948.
  5. GOOD: A Graph Out-of-Distribution Benchmark. In NeurIPS 2022.
  6. A Data-centric Framework to Endow Graph Neural Networks with Out-Of-Distribution Detection Ability. In KDD 2023.
  7. Open-World Lifelong Graph Learning. In IJCNN 2023.
  8. End-to-end open-set semi-supervised node classification with out-of-distribution detection. In IJCAI 2022.
  9. A survey on fairness-aware recommender systems. Information Fusion, 100: 101906.
  10. CGMN: A Contrastive Graph Matching Network for Self-Supervised Graph Similarity Learning. In IJCAI 2022.
  11. Dual Intent Enhanced Graph Neural Network for Session-based New Item Recommendation. In WWW 2023.
  12. Empowering Graph Representation Learning with Test-Time Graph Transformation. In ICLR 2023.
  13. Adam: A method for stochastic optimization. In ICLR 2014.
  14. Semi-Supervised Classification with Graph Convolutional Networks. In ICLR 2017.
  15. On information and sufficiency. The annals of mathematical statistics, 22(1): 79–86.
  16. GraphDE: A Generative Framework for Debiased Learning and Out-of-Distribution Detection on Graphs. In NeurIPS 2022.
  17. Isolation Forest. In ICDM 2008.
  18. GOOD-D: On Unsupervised Graph Out-Of-Distribution Detection. In WSDM 2023.
  19. Towards Self-Interpretable Graph-Level Anomaly Detection. In NeurIPS 2023.
  20. Learning Strong Graph Neural Networks with Weak Information. In KDD 2023.
  21. Deep Graph-level Anomaly Detection by Glocal Knowledge Distillation. In WSDM 2022.
  22. One-Class SVMs for Document Classification. Journal of machine Learning research, 2: 139–154.
  23. TUDataset: A collection of benchmark datasets for learning with graphs. In ICML Workshop 2020.
  24. Propagation kernels: efficient graph kernels from propagated information. Machine learning, 102(2): 209–245.
  25. SSD: A Unified Framework for Self-Supervised Outlier Detection. In ICLR 2021.
  26. Weisfeiler-Lehman Graph Kernels. Journal of Machine Learning Research, 12: 2539–2561.
  27. Sklar, A. 1973. Random variables, joint distribution functions, and copulas. Kybernetika, 9(6): 449–460.
  28. Sklar, M. 1959. Fonctions de répartition à n dimensions et leurs marges. In Annales de l’ISUP.
  29. Learning on Graphs with Out-of-Distribution Nodes. In KDD 2022.
  30. Graph posterior network: Bayesian predictive uncertainty for node classification. NeurIPS 2021, 34: 18033–18048.
  31. InfoGraph: Unsupervised and Semi-supervised Graph-Level Representation Learning via Mutual Information Maximization. In ICLR 2020.
  32. Graph Structure Learning with Variational Information Bottleneck. In AAAI 2022.
  33. Learning and Evaluating Graph Neural Network Explanations based on Counterfactual and Factual Reasoning. In WWW 2022.
  34. Is heterogeneity notorious? taming heterogeneity to handle test-time shift in federated learning. In NeurIPS 2023.
  35. Federated learning on non-iid graphs via structural knowledge sharing. In AAAI 2023.
  36. An exact algorithm for minimum vertex cover problem. Mathematics, 7(7): 603.
  37. An Exact Algorithm for Minimum Weight Vertex Cover Problem in Large Graphs. CoRR, abs/1903.05948.
  38. Contrastive Graph Similarity Networks. ACM Transactions on the Web.
  39. Augmenting Affective Dependency Graph via Iterative Incongruity Graph Learning for Sarcasm Detection. In AAAI 2023.
  40. GraphGuard: Detecting and Counteracting Training Data Misuse in Graph Neural Networks. In NDSS.
  41. On solving simplified diversified top-k s-plex problem. Computers & Operations Research, 153: 106187.
  42. HEA-D: A Hybrid Evolutionary Algorithm for Diversified Top-k Weight Clique Search Problem. In IJCAI 2022. Main Track.
  43. Energy-based Out-of-Distribution Detection for Graph Neural Networks. In ICLR 2023.
  44. Graph information bottleneck. NeurIPS 2020, 33: 20437–20448.
  45. How Powerful are Graph Neural Networks? In ICLR 2019.
  46. Graph Contrastive Learning with Augmentations. In NeurIPS 2020.
  47. Graph Information Bottleneck for Subgraph Recognition. In ICLR 2021.
  48. Commonsense Knowledge Enhanced Sentiment Dependency Graph for Sarcasm Detection. In IJCAI 2023.
  49. Demystifying Uneven Vulnerability of Link Stealing Attacks against Graph Neural Networks. In ICML, volume 202 of Proceedings of Machine Learning Research, 41737–41752. PMLR.
  50. Projective Ranking: A Transferable Evasion Attack Method on Graph Neural Networks. In CIKM, 3617–3621. ACM.
  51. Trustworthy Graph Neural Networks: Aspects, Methods and Trends. CoRR, abs/2205.07424.
  52. On the Interaction between Node Fairness and Edge Privacy in Graph Neural Networks. CoRR, abs/2301.12951.
  53. Projective Ranking-Based GNN Evasion Attacks. IEEE Trans. Knowl. Data Eng., 35(8): 8402–8416.
  54. Uncertainty Aware Semi-Supervised Learning on Graph Data. In NeurIPS 2020.
  55. Towards Data-centric Graph Machine Learning: Review and Outlook. arXiv preprint arXiv:2309.10979.
  56. GNNEvaluator: Evaluating GNN Performance On Unseen Graphs Without Labels. NeurIPS 2023.
  57. Large language models for scientific synthesis, inference and explanation. arXiv preprint arXiv:2310.07984.
  58. Heterogeneous graph attention network for small and medium-sized enterprises bankruptcy prediction. In PAKDD 2021.
  59. Rethinking and scaling up graph contrastive learning: An extremely efficient approach with group discrimination. NeurIPS 2022.
  60. Finding the Missing-half: Graph Complementary Learning for Homophily-prone and Heterophily-prone Graphs. In ICML 2023.
Citations (5)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
Github Logo Streamline Icon: https://streamlinehq.com

GitHub

  1. GitHub - Ee1s/GOODAT: GOODAT: Towards Test-time Graph Out-of-Distribution Detection (8 stars)