Few-Shot Calibration of Set Predictors via Meta-Learned Cross-Validation-Based Conformal Prediction

Conventional frequentist learning is known to yield poorly calibrated models that fail to reliably quantify the uncertainty of their decisions. Bayesian learning can improve calibration, but formal guarantees apply only under restrictive assumptions about correct model specification. Conformal predi...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	IEEE transactions on pattern analysis and machine intelligence 2024-01, Vol.46 (1), p.280-291
Hauptverfasser:	Park, Sangwoo, Cohen, Kfir M., Simeone, Osvaldo
Format:	Artikel
Sprache:	eng
Schlagworte:	Bayes methods Calibration Conformal prediction cross-validation distribution-free calibration Machine learning meta-learning Metalearning set prediction Task analysis Training Training data Uncertainty
Online-Zugang:	Volltext bestellen
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!