all AI news
The Calibration Generalization Gap. (arXiv:2210.01964v2 [cs.LG] UPDATED)
Oct. 7, 2022, 1:14 a.m. | A. Michael Carrell, Neil Mallinar, James Lucas, Preetum Nakkiran
stat.ML updates on arXiv.org arxiv.org
Calibration is a fundamental property of a good predictive model: it requires
that the model predicts correctly in proportion to its confidence. Modern
neural networks, however, provide no strong guarantees on their calibration --
and can be either poorly calibrated or well-calibrated depending on the
setting. It is currently unclear which factors contribute to good calibration
(architecture, data augmentation, overparameterization, etc), though various
claims exist in the literature.
We propose a systematic way to study the calibration error: by decomposing …
More from arxiv.org / stat.ML updates on arXiv.org
Jobs in AI, ML, Big Data
Data Architect
@ University of Texas at Austin | Austin, TX
Data ETL Engineer
@ University of Texas at Austin | Austin, TX
Lead GNSS Data Scientist
@ Lurra Systems | Melbourne
Senior Machine Learning Engineer (MLOps)
@ Promaton | Remote, Europe
Data Management Assistant
@ World Vision | Amman Office, Jordan
Cloud Data Engineer, Global Services Delivery, Google Cloud
@ Google | Buenos Aires, Argentina