Comparing Interval-Valued Estimations with Point-Valued Estimations - LIRMM - Laboratoire d’Informatique, de Robotique et de Microélectronique de Montpellier Accéder directement au contenu
Communication Dans Un Congrès Année : 2016

Comparing Interval-Valued Estimations with Point-Valued Estimations

Olivier Strauss
Ines Couso
  • Fonction : Auteur
  • PersonId : 974635

Résumé

In the last decade, numerous proposals have been made to deal with imprecision in estimation problems. Those approaches, many of which involve dealing with interval-valued outputs, deal with the subtle difference between uncertainty and imprecision. One of the crucial points − which to our knowledge has never been addressed − is “how to compare an interval-valued method with a precise valued method?” The usual way to compare two estimation methods is to use benchmark data with ground truths and to compute a distance between the estimates of each method and the ground truth. However, most of the mathematical available extensions of distances are either biased in favor of a precise approach or in favor of an imprecise approach. This paper proposes a new tool, the weighted variation of the mid-point distance (WVD), that is more suitable to achieve this kind of comparison, dealing with imprecision with a particular semantic. After reviewing existing distances, we introduce the WVD, first from an intuitive perspective, then from a more mathematical point of view. Its very satisfactory properties are highlighted through an experiment.
Fichier principal
Vignette du fichier
SaulnierCousoStrauss2016.pdf (469.87 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

lirmm-01488052 , version 1 (17-05-2022)

Identifiants

Citer

Hugo Saulnier, Olivier Strauss, Ines Couso. Comparing Interval-Valued Estimations with Point-Valued Estimations. IPMU 2016 - 16th International Conference on Information Processing and Management of Uncertainty in Knowledge-Based Systems, Jun 2016, Eindhoven, Netherlands. pp.595-604, ⟨10.1007/978-3-319-40581-0_48⟩. ⟨lirmm-01488052⟩
117 Consultations
33 Téléchargements

Altmetric

Partager

Gmail Mastodon Facebook X LinkedIn More