Salvato in:
Dettagli Bibliografici
Autore principale: Coleman, Charles D.
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2511.14812
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911275154407424
author Coleman, Charles D.
author_facet Coleman, Charles D.
contents Level-based and share-based loss functions are asymptotically equivalent if, in the limit, their averages converge almost surely to a constant ratio. These loss functions take a target value and its realization as arguments and are often used to measure accuracy. The equivalence is proved for a large class of loss functions, the weighted exponentiated functions, when the weights are decomposable as a particular product form. An upshot is that when losses are averaged for a large number of units, differences in ratios and, hence, ranks, are negligible, when the average (or summed) difference between the target values and their realizations is around zero. This implies the almost sure asymptotic convergence of numerical and distributive accuracy when using these loss functions.
format Preprint
id arxiv_https___arxiv_org_abs_2511_14812
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle The Asymptotic Equivalence of Level-Based and Share-Based Loss Functions
Coleman, Charles D.
Statistics Theory
Methodology
91C05 62C99
Level-based and share-based loss functions are asymptotically equivalent if, in the limit, their averages converge almost surely to a constant ratio. These loss functions take a target value and its realization as arguments and are often used to measure accuracy. The equivalence is proved for a large class of loss functions, the weighted exponentiated functions, when the weights are decomposable as a particular product form. An upshot is that when losses are averaged for a large number of units, differences in ratios and, hence, ranks, are negligible, when the average (or summed) difference between the target values and their realizations is around zero. This implies the almost sure asymptotic convergence of numerical and distributive accuracy when using these loss functions.
title The Asymptotic Equivalence of Level-Based and Share-Based Loss Functions
topic Statistics Theory
Methodology
91C05 62C99
url https://arxiv.org/abs/2511.14812