Research map / Label-efficient and robust learning / Recognition and long-tail
Conformal Prediction with Corrupted Labels: Uncertain Imputation and Robust Re-weighting
Shai Feldman, Stephen Bates, Yaniv Romano
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 uncertain imputation instead of privileged conformal prediction when weights are hard to estimate | UI · Synthetic, hard-to-estimate missingness (target 90%) | empirical coverage | Improved |
| combine Naive CP, PCP and UI by union (TriplyRobust) | TriplyRobust · Synthetic, all components degenerate | empirical coverage | Hurt |
| have at least one oracle component in TriplyRobust (ablation) | TriplyRobust · Synthetic, degenerate/oracle combinations | empirical coverage | Improved |
| use UI with sampled errors instead of naive mean imputation (ablation) | UI · Real datasets with 20% labels removed | empirical coverage | Improved |
| use PCP weights instead of Naive CP | PCP · Real datasets with 20% labels removed | empirical coverage | Improved |
| use PCP with estimated weights instead of true weights (ablation) | PCP · Real datasets with 20% labels removed | empirical coverage | No clear effect |
| use PCP when Naive CP over-covers and weights shifted by delta >= 0 | PCP · Synthetic, shifted weights | empirical coverage | No clear effect |
| use PCP when Naive CP under-covers and weights shifted outside (-W/(n+1), 0) | PCP · Synthetic, shifted weights | empirical coverage | Hurt |
Closest papers by meaning
- Exploring the Noise Robustness of Online Conformal Prediction (NeurIPS 2025 · similarity 0.90)
- Robust Conformal Prediction Using Privileged Information (NeurIPS 2024 · similarity 0.90)
- Provably Reliable Conformal Prediction Sets in the Presence of Data Poisoning (ICLR 2025 · similarity 0.89)
- Conformal Prediction with Cellwise Outliers: A Detect-then-Impute Approach (ICML 2025 · similarity 0.88)
- One Sample is Enough to Make Conformal Prediction Robust (NeurIPS 2025 · similarity 0.88)
- Conformal Prediction for Class-wise Coverage via Augmented Label Rank Calibration (NeurIPS 2024 · similarity 0.88)
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.
