Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Symmetric Aggregation of Conformity Scores for Efficient Uncertainty Sets
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 |
|---|---|---|---|
| optimize the exponent p of the power-sum aggregator (SACP++ vs default p=1) | SACP · CIFAR-10 (alpha=0.05) | prediction set length: 1.308 -> 1.281 | Improved |
| optimize the exponent p of the power-sum aggregator | SACP · MNIST (alpha=0.1) | prediction set length: 0.904 -> 0.900 | Improved |
| aggregate e-value-normalized scores symmetrically (SACP++) instead of selecting the best single model | best base model (BL) · CIFAR-10 (alpha=0.05) | prediction set length: 1.531 -> 1.281 | Improved |
| score-level SACP++ vs majority-vote merging CM | CM · CIFAR-10 (alpha=0.05) | prediction set length: 2.089 -> 1.281 | Improved |
| score-level SACP++ vs randomized majority vote CR | CR · CIFAR-10 (alpha=0.1) | prediction set length: 1.194 -> 1.028 | Improved |
| SACP++ vs Wagg weighted aggregation | Wagg · CIFAR-10 (alpha=0.05) | prediction set length: 1.292 -> 1.281 | Improved |
| SACP++ vs CSA multivariate quantile aggregation | CSA · MNIST (alpha=0.05) | prediction set length: 0.958 -> 0.954 | Improved |
| SACP vs Wagg without data splitting | Wagg · CIFAR-10 (alpha=0.1) | prediction set length: 1.032 vs 1.034 | No clear effect |
| aggregate predictors on regression datasets (SACP++ vs best single model) | best individual model BL · OpenML regression (alpha=0.05) | prediction set length | Mixed |
| SACP++ vs other aggregation methods on regression | aggregation baselines · OpenML regression (alpha=0.05) | prediction set length | Improved |
Closest papers by meaning
- Conformal Prediction for Ensembles: Improving Efficiency via Score-Based Aggregation (NeurIPS 2025 · similarity 0.91)
- Conformity Score Averaging for Classification (ICML 2025 · similarity 0.88)
- Provably Reliable Conformal Prediction Sets in the Presence of Data Poisoning (ICLR 2025 · similarity 0.87)
- Conformal Prediction with Learned Features (ICML 2024 · similarity 0.87)
- Multi-model Ensemble Conformal Prediction in Dynamic Environments (NeurIPS 2024 · similarity 0.87)
- Conformalized Interval Arithmetic with Symmetric Calibration (AAAI 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.
