Research map / Learning theory and algorithms / Prediction and decision losses
Colorful Pinball: Density-Weighted Quantile Regression for Conditional Guarantee of Conformal Prediction
Qianyi Chen, Bo Li
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 |
|---|---|---|---|
| add density-weighted objective (CPCP) over RCP (ablation) | RCP · Naval | MSCE: 0.0029 -> 0.0019 | Improved |
| add density-weighted objective (CPCP) over RCP (ablation) | RCP · SGEMM | MSCE: 0.0007 -> 0.0003 | Improved |
| add density-weighted objective (CPCP) over RCP (ablation) | RCP · Naval | WSC: 0.8002 -> 0.8169 | Improved |
| co-train three quantile heads only (RCP-MultiHead) over RCP (ablation) | RCP · Naval | MSCE: 0.0029 -> 0.0027 | No clear effect |
| add clipping and mixing of weights to CPCP (ablation) | CPCP · WEC | MSCE: 0.0025 -> 0.0012 | Improved |
| add clipping and mixing of weights to CPCP (ablation) | CPCP · Naval | WSC: 0.8169 -> 0.8320 | Improved |
| add clipping and mixing of weights to CPCP (ablation) | CPCP · Naval | MSCE: 0.0019 -> 0.0019 | No clear effect |
| add mixing only to CPCP (ablation) | CPCP · Naval | MSCE: 0.0019 -> 0.0014 | Improved |
| add clipping only to CPCP (ablation) | CPCP · WEC | MSCE: 0.0025 -> 0.0016 | Improved |
| use finite-difference bandwidth 0.01 instead of default 0.02 (ablation) | CPCP (Clip+Mix) · WEC | MSCE: 0.0012 -> 0.0013 | No clear effect |
| use CPCP instead of CQR | CQR · Naval | MSCE: 0.0120 -> 0.0019 | Improved |
| use CPCP (Clip+Mix) instead of Split conformal | Split · SGEMM | WSC: 0.7435 -> 0.8912 | Improved |
Closest papers by meaning
- Rectifying Conformity Scores for Better Conditional Coverage (ICML 2025 · similarity 0.92)
- Probabilistic Conformal Prediction with Approximate Conditional Validity (ICLR 2025 · similarity 0.91)
- Co-optimization for Adaptive Conformal Prediction (NeurIPS 2026 · similarity 0.89)
- Kandinsky Conformal Prediction: Beyond Class- and Covariate-Conditional Coverage (ICML 2025 · similarity 0.89)
- Backward Conformal Prediction (NeurIPS 2025 · similarity 0.87)
- Conformal Inference under High-Dimensional Covariate Shifts via Likelihood-Ratio Regularization (NeurIPS 2025 · similarity 0.87)
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.
