Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Simulation-based Calibration of Uncertainty Intervals under Approximate Bayesian Estimation
Terrance D. Savitsky, Julie Gershunskaya
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 rescaling calibration to ADVI posterior intervals | ADVI mean-field VB, Fay-Herriot model · Fay-Herriot simulation | coverage (50% nominal): 0.531 -> 0.493 | Improved |
| add pivotal calibration to ADVI posterior intervals | ADVI mean-field VB, Fay-Herriot model · Fay-Herriot simulation | coverage (50% nominal): 0.531 -> 0.492 | Improved |
| add calibration (rescaled or pivot) to reduce interval length | ADVI, Fay-Herriot model · Fay-Herriot simulation | interval length: 1.019 -> 0.935 (rescaled), 0.933 (pivot) | Improved |
| add rescaling calibration | ADVI, CFHV mixture model · CFHV simulation | coverage (50% nominal): 0.625 -> 0.559 | Improved |
| add pivotal calibration | ADVI, CFHV mixture model · CFHV simulation | coverage (50% nominal): 0.625 -> 0.549 | Improved |
| use pivotal instead of rescaled calibration | calibrated ADVI intervals · CES real-data simulation, Jan | coverage (50% nominal): 0.523 -> 0.503 | Improved |
| add pivotal calibration on CES data | ADVI, CES Leisure sector · CES real-data simulation, Dec | coverage (50% nominal): 0.604 -> 0.529 | Improved |
| add pivotal calibration on CES data | ADVI, CES Leisure sector · CES real-data simulation, Feb | coverage (50% nominal): 0.572 -> 0.500 | Improved |
Closest papers by meaning
- Bayesian Adaptive Calibration and Optimal Design (NeurIPS 2024 · similarity 0.85)
- Asymptotically Log-Optimal Bayes-Assisted Confidence Sequences for Bounded Means (NeurIPS 2026 · similarity 0.85)
- Learning Likelihood-Free Reference Priors (ICML 2025 · similarity 0.85)
- Consistency Models for Scalable and Fast Simulation-Based Inference (NeurIPS 2024 · similarity 0.85)
- Addressing Misspecification in Simulation-based Inference through Data-driven Calibration (ICML 2025 · similarity 0.84)
- Online Bayesian Calibration under Gradual and Abrupt System Changes (NeurIPS 2026 · similarity 0.84)
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.
