Abstract
We propose a method for reliable prediction in multi-class classification, where reliability refers to the possibility of partial abstention in cases of uncertainty. More specifically, we allow for predictions in the form of preorder relations on the set of classes, thereby generalizing the idea of set-valued predictions. Our approach relies on combining learning by pairwise comparison with a recent proposal for modeling uncertainty in classification, in which a distinction is made between reducible (a.k.a. epistemic) uncertainty caused by a lack of information and irreducible (a.k.a. aleatoric) uncertainty due to intrinsic randomness. The problem of combining uncertain pairwise predictions into a most plausible preorder is then formalized as an integer programming problem. Experimentally, we show that our method is able to appropriately balance reliability and precision of predictions.
Dokumententyp: | Konferenzbeitrag (Paper) |
---|---|
Fakultät: | Mathematik, Informatik und Statistik > Informatik > Künstliche Intelligenz und Maschinelles Lernen |
Themengebiete: | 000 Informatik, Informationswissenschaft, allgemeine Werke > 004 Informatik |
Ort: | Menlo Park, County of San Mateo, California |
Bemerkung: | ISBN 978-0-9992411-2-7. Die Konferenzen "International Joint Conference on Artificial Intelligence (IJCAI)" und "European Conference on Artificial Intelligence (ECAI)" fanden 2018 erstmals gemeinsam statt, unter dem Namen "International Joint Conference on Artificial Intelligence and European Conference on Artificial Intelligence (IJCAI-ECAI)". Obwohl im Titel nur die Konferenz IJCAI angegeben ist, handelt es sich nach Recherche auf der Konferenzwebseite und im Internet um die gemeinsamen Proceedings der IJCAI-ECAI |
Sprache: | Englisch |
Dokumenten ID: | 91705 |
Datum der Veröffentlichung auf Open Access LMU: | 31. Mrz. 2022 12:53 |
Letzte Änderungen: | 12. Okt. 2024 19:36 |