A Distributionally Robust Estimator that Dominates the Empirical Average

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Koumpis, Nikolas, Kalogerias, Dionysis
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866929391642083328
author Koumpis, Nikolas
Kalogerias, Dionysis
author_facet Koumpis, Nikolas
Kalogerias, Dionysis
contents We leverage the duality between risk-averse and distributionally robust optimization (DRO) to devise a distributionally robust estimator that strictly outperforms the empirical average for all probability distributions with negative excess kurtosis. The aforesaid estimator solves the $χ^{2}-$robust mean squared error problem in closed form.
format Preprint
id arxiv_https___arxiv_org_abs_2402_10418
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A Distributionally Robust Estimator that Dominates the Empirical Average
Koumpis, Nikolas
Kalogerias, Dionysis
Statistics Theory
We leverage the duality between risk-averse and distributionally robust optimization (DRO) to devise a distributionally robust estimator that strictly outperforms the empirical average for all probability distributions with negative excess kurtosis. The aforesaid estimator solves the $χ^{2}-$robust mean squared error problem in closed form.
title A Distributionally Robust Estimator that Dominates the Empirical Average
topic Statistics Theory
url https://arxiv.org/abs/2402.10418