Get Started
Home
Topics
Search
Library
Research questionHow can one released graph dataset resist learning across multiple downstream graph tasks?A perturbation that disrupts learning for one graph task may leave the same released data useful for another. Data owners therefore need protection that transfers across plausible downstream uses rather than targeting a single task.
AI
Machine Learning
Research Paper
Statistical Machine Learning
Technology
Latest papersRecent research connected to this question, newest first.MUGEN: Generating Unlearnable Graph Examples for Multiple Learning TasksThe source studies a single feature-perturbed graph release intended to protect node classification, graph classification, and link prediction. Evidence covers continuous and discrete node features, multiple GNN backbones, five benchmarks, and robustness under adversarial training and data augmentation.research paper · Sep 3, 2026
Related questions
How can collaborative learning preserve graph structure when graph data are distributed across agents?How can graph neural networks capture long-range interactions without oversmoothing or oversquashing on diverse large graphs?How can graph-level anomaly detection transfer to new domains with little or no target data?How can reservoir-based graph neural networks aggregate multi-hop information without over-smoothing?