Deeply Uncertain: Comparing Methods of Uncertainty Quantification in Deep Learning Algorithms [Slides]
- Fermi National Accelerator Lab. (FNAL), Batavia, IL (United States)
Deep learning is used in many applications in the physical sciences. In those sciences we are used to having uncertainty on every measurement or prediction. In the last years, many methods have been put forth for uncertainty quantification: Bayesian Neural Networks, Concrete Dropout, and Deep Ensembles. Uncertainty in deep learning is often divided into aleatoric or irreducible and epistemic or reducible. The problem remains: Which uncertainty quantification method should be chosen? How are those results interpreted?
- Research Organization:
- Fermi National Accelerator Lab. (FNAL), Batavia, IL (United States)
- Sponsoring Organization:
- USDOE Office of Science (SC), High Energy Physics (HEP)
- DOE Contract Number:
- AC02-07CH11359
- OSTI ID:
- 1623354
- Report Number(s):
- FERMILAB-SLIDES-20-008-SCD; oai:inspirehep.net:1796920; TRN: US2106536
- Resource Relation:
- Conference: 1.Fundamental Science in the AI Era at International Conference on Learning Representations (ICLR) 2020, (Held Virtually), 26 Apr 2020
- Country of Publication:
- United States
- Language:
- English
Similar Records
Deeply uncertain: comparing methods of uncertainty quantification in deep learning algorithms
Prediction of the SYM-H Index Using a Bayesian Deep Learning Method With Uncertainty Quantification
Semi-supervised Bayesian Low-shot Learning
Journal Article
·
Tue Dec 01 00:00:00 EST 2020
· Machine Learning: Science and Technology
·
OSTI ID:1623354
Prediction of the SYM-H Index Using a Bayesian Deep Learning Method With Uncertainty Quantification
Journal Article
·
Wed Feb 14 00:00:00 EST 2024
· Space Weather (Online)
·
OSTI ID:1623354
+5 more
Semi-supervised Bayesian Low-shot Learning
Technical Report
·
Wed Sep 01 00:00:00 EDT 2021
·
OSTI ID:1623354
+3 more