The Costs of Indeterminacy: How to Determine Them? - Archive ouverte HAL Access content directly
Journal Articles IEEE Transactions on Cybernetics Year : 2017

The Costs of Indeterminacy: How to Determine Them?

Abstract

Indeterminate classifiers are cautious models able to predict more than one class in case of high uncertainty. A problem that arises when using such classifiers is how to evaluate their performances. This problem has already been considered in the case where all prediction errors have equivalent costs (that we will refer as the ''0/1 costs'' or accuracy setting). The purpose of this paper is to study the case of generic cost functions. We provide some properties that the costs of indeterminate predictions could or should follow, and review existing proposals in the light of those properties. This allows us to propose a general formula fitting our properties that can be used to produce and evaluate indeterminate predictions. Some experiments on the cost-sensitive problem of ordinal regression illustrate the behavior of the proposed evaluation criterion.
Fichier principal
Vignette du fichier
comparisonClassifiers_Final_IEEE.pdf (544.99 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01378361 , version 1 (21-06-2021)

Identifiers

Cite

Gen Yang, Sébastien Destercke, Marie-Hélène Masson. The Costs of Indeterminacy: How to Determine Them?. IEEE Transactions on Cybernetics, 2017, 47 (12), pp.4316-4327. ⟨10.1109/TCYB.2016.2607237⟩. ⟨hal-01378361⟩
83 View
56 Download

Altmetric

Share

Gmail Facebook X LinkedIn More