Misspecification-Averse Estimation

Isaiah Andrews, Ricky Li, Yucheng Shang · 2026

Read Misspecification-Averse Estimation by Isaiah Andrews, Ricky Li, Yucheng Shang (2026) 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-2604-23176.txt.

Introduction

Andrews, Li, and Shang study optimal estimation when the likelihood may be misspecified, building on choice theory under uncertainty. They analyse axiom-based optimality rules nesting several current misspecification-robust goals, and introduce the constrained multiplier rule for flexible misspecification views. A local asymptotic minimax result extends a classic efficiency bound to a limit experiment with moment-limited misspecification. Asymptotically optimal estimators emerge as Bayes rules under a flat prior and a tilted likelihood with moment limits. Feasible plug-in versions are also asymptotically optimal.

Catalog record

CREATORIsaiah Andrews, Ricky Li, Yucheng Shang
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE169635
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.23176

Browse the full library · About the Tower of Babel