Accéder directement au contenu Accéder directement à la navigation
Communication dans un congrès

Explaining a result to the end-user: a geometric approach for classification problems

Abstract : This paper addresses the issue of the explanation of the result given to the end-user by a classier, when it is used as a decision support system. We consider machine learning classiers, which provide a class for new cases, but also deterministic classiers that are built to solve a particular problem (like in viability or control problems). The end-user relies mainly on global information (like error rates) to assess the quality of the result given by the system. Even class membership probability, if available, describes only the statistical viewpoint, it doesn't take into account the context of a particular case. In the case of numerical state space, we propose to use the decision boundary of the classier (which always exists, even implicitly), to describe the situation of a particular case: The distance of a case to the decision boundary measures the ro-bustness of the decision to a change in the input data. Other geometric concepts can present a precise picture of the situation to the end-user. This geometric study is applied to different types of classiers.
Type de document :
Communication dans un congrès
Liste complète des métadonnées

Littérature citée [16 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-00493909
Déposant : Import Ws Irstea <>
Soumis le : lundi 21 juin 2010 - 15:47:05
Dernière modification le : mercredi 18 novembre 2020 - 10:26:20
Archivage à long terme le : : mercredi 22 septembre 2010 - 18:12:17

Fichier

CF2009-PUB00027387.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-00493909, version 1
  • IRSTEA : PUB00027387

Collections

Citation

I. Alvarez, S. Martin. Explaining a result to the end-user: a geometric approach for classification problems. Exact09, IJCAI 2009 Workshop on explanation aware computing (International Joint Conferences on Artificial Intelligence), Jul 2009, Pasadena, United States. p. 102 - p. 109. ⟨hal-00493909⟩

Partager

Métriques

Consultations de la notice

267

Téléchargements de fichiers

95