Research map / Graph learning / Graph algorithms and clustering
On the Discrimination Risk of Mean Aggregation Feature Imputation in Graphs
Arjun Subramonian, Kai-Wei Chang, Yizhou Sun
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use 0-fair imputation instead of regular Global Mean | Global Mean imputation · SBM | discrimination risk: 0.051 -> 0 | Improved |
| use 0-fair imputation instead of regular Neighbor Mean | Neighbor Mean imputation · SBM | discrimination risk: 0.048 -> 0 | Improved |
| use eps-fair constraint with eps=0 instead of regular on Feature Propagation | Feature Propagation · SBM | discrimination risk: 0.028 -> 0 | Improved |
| use eps-fair constraint with eps=0 on Graph Regularization | Graph Regularization · SBM | discrimination risk: 0.021 -> 0 | Improved |
| 0-fair instead of regular Global Mean | Global Mean imputation · SBM | relative reconstruction error: 1 -> 1.21 | Hurt |
| 0-fair instead of regular Feature Propagation | Feature Propagation · SBM | relative reconstruction error: 0.977 -> 1.184 | Hurt |
| 0-fair instead of regular Graph Regularization | Graph Regularization · SBM | relative reconstruction error: 0.977 -> 1.006 | Hurt |
| 0-fair Feature Propagation instead of regular, linear group-membership classifier | Feature Propagation · SBM | group membership identification accuracy (linear): 0.814 -> 0.6 | Improved |
| 0-fair instead of regular, GCN group-membership classifier | Feature Propagation · SBM | group membership identification accuracy (gcn): 0.612 -> 0.505 | Improved |
| 0-fair instead of regular, MLP group-membership classifier | Graph Regularization · SBM | group membership identification accuracy (mlp): 0.821 -> 0.713 | Improved |
| lower eps from 0.05 to 0 in fair Global Mean (ablation) | eps-fair Global Mean · SBM | group membership identification accuracy (linear): 0.736 -> 0.602 | Improved |
| eps-fair constraints on Feature Propagation / Graph Regularization (structure-aware) vs fair GM (ablation) | eps-fair imputation · SBM | relative reconstruction error: 1.21 -> 1.006 | Improved |
Closest papers by meaning
- Propagate and Inject: Revisiting Propagation-Based Feature Imputation for Graphs with Partially Observed Features (ICML 2025 · similarity 0.86)
- FairSIN: Achieving Fairness in Graph Neural Networks through Sensitive Information Neutralization (AAAI 2024 · similarity 0.85)
- Attribute-Missing Graph Clustering Network (AAAI 2024 · similarity 0.85)
- Fair Graph U-Net: A Fair Graph Learning Framework Integrating Group and Individual Awareness (AAAI 2025 · similarity 0.85)
- Information Loss and Disparate Effects in Network Embeddings (NeurIPS 2026 · similarity 0.84)
- Random features models: a way to study the success of naive imputation (ICML 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
