U-TraCE: a conformal prediction approach to uncertainty quantification in black-box models

Modern machine learning (ML) systems, particularly deep neural networks, have become increasingly complex and opaque, making reliable uncertainty quantification (UQ) essential. Conventional UQ methods often prove inadequate for black-box model evaluation scenarios, either because model internals are...

Descripción completa

Detalles Bibliográficos

Autores principales:	Edgardo J Marchi, Maik Liebl
Formato:	Artigo
Lenguaje:	Inglês
Publicado:	IOP Publishing 2026-01-01
Colección:	Machine Learning: Science and Technology
Materias:	uncertainty quantification conformal prediction U-TraCE
Acceso en línea:	https://doi.org/10.1088/2632-2153/ae35ce
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!