Research map / Learning theory and algorithms / Matrix and tensor methods
Adversarial Regression with Doubly Non-negative Weighting Matrices
Tam Le, Truyen Nguyen, Makoto Yamada, Jose Blanchet, Viet Anh Nguyen
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 Mahalanobis-distance Gaussian kernel (NW-Metric) instead of standard NW | Nadaraya-Watson estimator · 8 regression datasets, no perturbation | RMSE | Improved |
| add adversarial reweighting (NW-LogDet / NW-BuresW) to NW with no perturbation | NW estimator · 8 regression datasets, no perturbation | RMSE | No clear effect |
| add adversarial reweighting to NW under perturbed training responses | NW estimator · 8 regression datasets, perturbed tau levels | RMSE | Improved |
| perturb more training neighbors (larger tau) | NW, LLR, LLR-I, NW-Metric baselines · 8 regression datasets | RMSE | Hurt |
| increase uncertainty size rho beyond a good value (ablation) | NW-LogDet and NW-BuresW · perturbed data, N=50 | RMSE | Mixed |
| set nearest neighbor size N to 20 or 30 instead of 10 or 50 (ablation) | NW-LogDet and NW-BuresW · KIN under varying tau | RMSE | Improved |
| increase nearest neighbor size N (ablation) | adversarial reweighting schemes · computational time | computational cost | Hurt |
Closest papers by meaning
- Doubly Robust Instance-Reweighted Adversarial Training (ICLR 2024 · similarity 0.88)
- Model Agnostic Sample Reweighting for Out-of-Distribution Learning (ICML 2022 · similarity 0.86)
- Precise asymptotics of reweighted least-squares algorithms for linear diagonal networks (NeurIPS 2024 · similarity 0.86)
- Stable Learning via Sample Reweighting (AAAI 2020 · similarity 0.86)
- Gradients Weights improve Regression and Classification (JMLR 2016 · similarity 0.86)
- Thumb on the Scale: Optimal Loss Weighting in Last Layer Retraining (NeurIPS 2025 · 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.
