Criteria of efficiency for set-valued classification. / Vovk, Vladimir.

In: Annals of Mathematics and Artificial Intelligence, 10.02.2017.

Research output: Contribution to journalArticle

Forthcoming

Abstract

We study optimal conformity measures for various criteria of efficiency of set-valued classification in an idealised setting. This leads to an important class of criteria of efficiency that we call probabilistic and argue for; it turns out that the most standard criteria of efficiency used in literature on conformal prediction are not probabilistic unless the problem of classification is binary. We consider both unconditional and label-conditional conformal prediction.
Original languageEnglish
JournalAnnals of Mathematics and Artificial Intelligence
DOIs
StateAccepted/In press - 10 Feb 2017
This open access research output is licenced under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.

ID: 27742780