Rethinking Mean Square Error: Information, Generalized Estimation, and the James–Stein Paradox
Paul Vos · 2024
Read Rethinking Mean Square Error: Information, Generalized Estimation, and the James–Stein Paradox by Paul Vos (2024) 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-2412-08475.txt.
Description
Paul Vos rethinks how estimators are judged in a 2024 paper. A famous paradox says a shrinkage estimator beats the classic maximum likelihood pick on squared error, though likelihood wins on most other counts. One camp questions likelihood; Vos questions the yardstick.
He contrasts point estimators with broader ones that use sample and parameter jointly, and pointwise scores with family-aware ones. A simple lemma shows no pointwise score can crown a uniform winner. An information score, by contrast, is topped by the classic score function itself. Under this lens the paradox dissolves: likelihood keeps all the information, shrinkage throws a known share away.
Catalog record
| CREATOR | Paul Vos |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-08475