Skip to Main content Skip to Navigation
Conference papers

Consistance de la minimisation du risque empirique pour l'optimisation de l'erreur relative moyenne

Abstract : We study in this paper the consequences of using the Mean Absolute Percentage Error (MAPE) as a measure of quality for regression models. We show that finding the best model under the MAPE is equivalent to doing weighted Mean Absolute Error (MAE) regression. We also show that, under some asumptions, universal consistency of Empirical Risk Minimization remains possible using the MAPE.
Document type :
Conference papers
Complete list of metadatas

Cited literature [3 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01195695
Contributor : Fabrice Rossi <>
Submitted on : Tuesday, September 8, 2015 - 12:05:14 PM
Last modification on : Thursday, October 1, 2020 - 1:22:01 PM
Long-term archiving on: : Wednesday, December 9, 2015 - 11:02:54 AM

Files

TexteLong_vmodif.pdf
Files produced by the author(s)

Licence


Distributed under a Creative Commons Attribution 4.0 International License

Identifiers

  • HAL Id : hal-01195695, version 1
  • ARXIV : 1509.02357

Collections

Citation

Arnaud de Myttenaere, Bénédicte Le Grand, Fabrice Rossi. Consistance de la minimisation du risque empirique pour l'optimisation de l'erreur relative moyenne. 47èmes Journées de Statistique de la SFdS, Société Française de Statistique, Jun 2015, Lille, France. ⟨hal-01195695⟩

Share

Metrics

Record views

233

Files downloads

261