Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Set-Preserving Calibration from Conformal P-Values to E-Values
Nabil Alami, Jad Zakharia, Souhaib Ben Taieb
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 set-preserving P2E calibrator instead of F1=-log(p) (ablation) | ECCP · Abalone (K=10, RF) | mean prediction set length: 29.07 -> 6.70 | Improved |
| use set-preserving P2E calibrator instead of F2 (ablation) | ECCP · Abalone (K=10, RF) | mean prediction set length: 10.21 -> 6.70 | Improved |
| use set-preserving P2E calibrator instead of F3 (ablation) | ECCP · Abalone (K=10, RF) | mean prediction set length: 32.02 -> 6.70 | Improved |
| use set-preserving P2E calibrator instead of F1 (ablation) | ECCP · Boston (K=10, RF) | mean prediction set length: 67.15 -> 9.76 | Improved |
| use P2E calibrator instead of Balinsky e-CP | p-value CP · Boston, RF | mean prediction set length: 50.816 -> 10.613 | Improved |
| use P2E calibrator instead of Balinsky e-CP | p-value CP · MNIST, linear | set size: 10.00 -> 1.07 | Improved |
| use WECA/UR-WECA instead of CM | conformal aggregation · OpenML task 361237 | prediction set length: 2.88 -> 1.64 | Improved |
| use UR-WECA instead of P_Agg | conformal aggregation · OpenML task 361244 | prediction set size: 6.91 -> 4.93 | Improved |
| use UR-WECA instead of COLA-S | conformal aggregation · OpenML task 361244 | prediction set size: 5.23 -> 4.93 | Improved |
| use UR-WECA instead of COLA-S | conformal aggregation · OpenML task 361234 | prediction set length: 2.90 -> 2.88 | No clear effect |
| use ECCP run at level 2alpha instead of CCP | cross-conformal · Boston (K=15, OLS) | prediction set size: 13.970 -> 10.665 | Improved |
| use WECA with F1 calibrator instead of UR-WECA (ablation) | WECA · OpenML task 361234 | coverage: 0.95 -> 1.00 | Hurt |
Closest papers by meaning
- Backward Conformal Prediction (NeurIPS 2025 · similarity 0.87)
- SimplexUQ: An Evaluation Framework and Benchmark for Conformal Uncertainty on Simplex-Valued Predictions (NeurIPS 2026 · similarity 0.87)
- Parametric Scaling Law of Tuning Bias in Conformal Prediction (ICML 2025 · similarity 0.86)
- False Coverage Proportion Control for Conformal Prediction (ICML 2025 · similarity 0.86)
- Improving the Statistical Efficiency of Cross-Conformal Prediction (ICML 2025 · similarity 0.86)
- Probabilistic Conformal Prediction with Approximate Conditional Validity (ICLR 2025 · similarity 0.86)
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.
