Search: onr:"swepub:oai:DiVA.org:liu-204029" >
Fast and Scalable S...
Fast and Scalable Score-Based Kernel Calibration Tests
-
- Glaser, Pierre (author)
- University College London, Gatsby Computational Neuroscience Unit, London, UK,UCL, Gatsby Computat Neurosci Unit, London, England
-
- Widmann, David (author)
- Uppsala universitet,Avdelningen för systemteknik
-
- Lindsten, Fredrik, 1984- (author)
- Linköpings universitet,Statistik och maskininlärning,Tekniska fakulteten,Linköping Univ, Div Stat & Machine Learning, Linköping, Sweden
-
show more...
-
- Gretton, Arthur (author)
- University College London, Gatsby Computational Neuroscience Unit, London, UK,UCL, Gatsby Computat Neurosci Unit, London, England
-
show less...
-
(creator_code:org_t)
- JMLR-JOURNAL MACHINE LEARNING RESEARCH, 2023
- 2023
- English.
-
In: Thirty-Ninth Conference on Uncertainty in Artificial Intelligence. - : JMLR-JOURNAL MACHINE LEARNING RESEARCH. ; , s. 691-700, s. 691-700
- Related links:
-
https://proceedings....
-
show more...
-
https://urn.kb.se/re...
-
https://urn.kb.se/re...
-
show less...
Abstract
Subject headings
Close
- We introduce the Kernel Calibration Conditional Stein Discrepancy test (KCCSD test), a non-parametric, kernel-based test for assessing the calibration of probabilistic models with well-defined scores. In contrast to previous methods, our test avoids the need for possibly expensive expectation approximations while providing control over its type-I error. We achieve these improvements by using a new family of kernels for score-based probabilities that can be estimated without probability density samples, and by using a conditional goodness-of-fit criterion for the KCCSD test’s U-statistic. The tractability of the KCCSD test widens the surface area of calibration measures to new promising use-cases, such as regularization during model training. We demonstrate the properties of our test on various synthetic settings.
Subject headings
- NATURVETENSKAP -- Matematik -- Sannolikhetsteori och statistik (hsv//swe)
- NATURAL SCIENCES -- Mathematics -- Probability Theory and Statistics (hsv//eng)
- NATURVETENSKAP -- Data- och informationsvetenskap -- Datavetenskap (hsv//swe)
- NATURAL SCIENCES -- Computer and Information Sciences -- Computer Sciences (hsv//eng)
Publication and Content Type
- ref (subject category)
- kon (subject category)
To the university's database