Research map / Learning theory and algorithms / Prediction and decision losses
On discriminative learning of prediction uncertainty
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| replace baseline uncertainty (plug-in risk or max-score) with learned uncertainty function minimizing the proposed order-enforcing loss | LR and SVM selective classifiers · 10 UCI/libSVM datasets (Table 2 subset) | AUC x100 (lower is better) | Helped |
| use learned uncertainty instead of SVM max-score | SVM selective classifier · LETTER, SHUTTLE, SENSORLESS | R@90 | Helped |
| learned uncertainty on MLP features vs quadratic features (ablation) | learned uncertainty function · across datasets | best-feature selection count: 6 vs 4 cases | Mixed |
| linear feature map for learned uncertainty (ablation) | learned uncertainty function · across datasets | selective risk | No clear effect |
Closest papers by meaning
- Regression with Cost-based Rejection (NeurIPS 2023 · similarity 0.88)
- Regression with reject option and application to kNN (NeurIPS 2020 · similarity 0.88)
- Rejection via Learning Density Ratios (NeurIPS 2024 · similarity 0.86)
- Learning with Selectively Labeled Data from Multiple Decision-makers (ICML 2025 · similarity 0.85)
- Leveraging Uncertainty Estimates To Improve Classifier Performance (ICLR 2024 · similarity 0.85)
- Learning to Reject with a Fixed Predictor: Application to Decontextualization (ICLR 2024 · similarity 0.83)
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.