Research map / Learning theory and algorithms / Prediction and decision losses
Improving Backward Conformal Prediction via Non-Conformity Score Transformation
Junxian Liu, Hao Zeng, Hongxin Wei
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 ST-BCP transformation instead of identity score | BCP · CIFAR-10 (n=200, T=2) | GAP (%): 5.38 -> 0.72 | Improved |
| use ST-BCP transformation instead of identity score | BCP · CIFAR-10 (n=200, T=2) | MSE (x10^4): 1.63 -> 0.79 | Improved |
| use ST-BCP transformation instead of identity score | BCP · CIFAR-10 (n=200, T=3) | STD (%): 1.16 -> 0.66 | Improved |
| use ST-BCP under constant size rule T=1 | BCP · CIFAR-10 (n=200, T=1) | GAP (%): 2.40 -> 1.90 | Improved |
| use ST-BCP with feature-and-calibration-dependent rule T(D,X) | BCP · CIFAR-10 (n=200, T(D,X)) | MisCov (%): 5.23 -> 5.23 | No clear effect |
| use ST-BCP with APS score (ablation) | BCP · CIFAR-10 (n=200, T=2) | GAP (%): 48.2 -> 0.84 | Improved |
| use ST-BCP with Rank score (ablation) | BCP · CIFAR-10 (n=200, T=2) | GAP (%): 35.6 -> 1.14 | Improved |
| use ST-BCP with DenseNet121 | BCP · CIFAR-10 (n=200, T=2) | GAP (%): 4 -> 0.7 | Improved |
| use ST-BCP with DenseNet121 | BCP · CIFAR-10 (n=200, T=2) | MSE (x10^4): 1.3 -> 0.6 | Improved |
| increase k in kNN used for T(D,X) from 20 to 70 (ablation) | T(D,X) size rule · CIFAR-10 (n=200) | fraction of unchanged constraints: 0.67 | No clear effect |
| increase k from 20 to 21 (ablation) | T(D,X) size rule · CIFAR-10 (n=200) | fraction of unchanged constraints: 0.92 | No clear effect |
| increase calibration set size n up to 800 | ST-BCP · CIFAR-10 (T=2) | MSE: order of 10^-5 | Improved |
Closest papers by meaning
- Backward Conformal Prediction (NeurIPS 2025 · similarity 0.92)
- Generative Conformal Prediction with Optimized Coverage Allocation (NeurIPS 2026 · similarity 0.91)
- Rectifying Conformity Scores for Better Conditional Coverage (ICML 2025 · similarity 0.91)
- Probabilistic Conformal Prediction with Approximate Conditional Validity (ICLR 2025 · similarity 0.90)
- Valid Selection among Conformal Sets (NeurIPS 2025 · similarity 0.90)
- Conformal Prediction with Learned Features (ICML 2024 · similarity 0.90)
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.
