Mix-n-Match : Ensemble and Compositional Methods for Uncertainty Calibration in Deep LearningDownload PDFOpen Website

2020 (modified: 17 May 2023)ICML 2020Readers: Everyone
Abstract: This paper studies the problem of post-hoc calibration of machine learning classifiers. We introduce the following desiderata for uncertainty calibration: (a) accuracy-preserving, (b) data-efficien...
0 Replies

Loading