Calibration Bands for Mean Estimates within the Exponential Dispersion Family
Łukasz Delong, Selim Gatti, Mario V. Wüthrich · 2025
Read Calibration Bands for Mean Estimates within the Exponential Dispersion Family by Łukasz Delong, Selim Gatti, Mario V. Wüthrich (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-18896.txt.
Description
Lukasz Delong, Selim Gatti, and Mario Wuthrich construct calibration bands for mean estimates in a 2025 paper. A model is calibrated when its mean estimates match the true means, but finite noisy samples make that ideal hard to reach. A weaker target, auto calibration, asks that responses average to the estimate given. Testing it has only recently drawn attention.
Calibration bands give lower and upper bounds that hold the true means at once with set confidence. Earlier bands existed for sub Gaussian and Bernoulli cases. The authors extend the construction to the whole exponential dispersion family, covering binomial, Poisson, gamma, and normal laws. The bands then yield tests for both calibration notions, without large sample approximations.
Catalog record
| CREATOR | Łukasz Delong, Selim Gatti, Mario V. Wüthrich |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 120 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-18896