mlsanityData to 7 Oct 2026

A Cautionary Tale of Decorrelating Theory Uncertainties

arXiv 2109.08159 ↗First posted 16 Sep 2021hep-ph cs.LG hep-excited by 11

Not on the map. Why?

Abstract (arXiv)

A variety of techniques have been proposed to train machine learning classifiers that are independent of a given feature. While this can be an essential technique for enabling background estimation, it may also be useful for reducing uncertainties. We carefully examine theory uncertainties, which typically do not have a statistical origin. We will provide explicit examples of two-point (fragmentation modeling) and continuous (higher-order corrections) uncertainties where decorrelating significantly reduces the apparent uncertainty while the actual uncertainty is much larger. These results suggest that caution should be taken when using decorrelation for these types of uncertainties as long as we do not have a complete decomposition into statistically meaningful components.

Key facts. Cited by 11 papers in the map (data through 7 Oct 2026). Builds on 17 of its 88 references that are in the map. First posted on arXiv 16 Sep 2021.

Builds on

17 of 88 references.

References · 17

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2014Unsupervised Domain Adaptation by Backpropagationcited by 1,906
  3. 2015Approximating Likelihood Ratios with Calibrated Discriminative Classifierscited by 87
  4. 2021A Living Review of Machine Learning for Particle Physicscited by 77
  5. 2018Mining gold from implicit models to improve likelihood-free inferencecited by 60
  6. 2016Learning to Pivot with Adversarial Networkscited by 55
  7. 2018Machine Learning in High Energy Physics Community White Papercited by 48
  8. 2015Jet-Images -- Deep Learning Editioncited by 41
  9. 2016Parameterized Machine Learning for High-Energy Physicscited by 33
  10. 2019MadMiner: Machine learning-based inference for particle physicscited by 32
  11. 2018Constraining Effective Field Theories with Machine Learningcited by 26
  12. 2018A Guide to Constraining Effective Field Theories with Machine Learningcited by 25
5 more
  1. 2013Partial Distance Correlation with Methods for Dissimilaritiescited by 20
  2. 2018INFERNO: Inference-Aware Neural Optimisationcited by 16
  3. 2017Decorrelated Jet Substructure Tagging using Adversarial Neural Networkscited by 12
  4. 2020Adversarial domain adaptation to reduce sample bias of a high energy physics classifiercited by 9
  5. 2020Optimal statistical inference in the presence of systematic uncertainties using neural network optimization based on binned Poisson likelihoods with nuisance parameterscited by 8

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 2; 2022: 4; 2023: 0; 2024: 2; 2025: 1; 2026: 2.

11 papers.

  1. 2021Machine Learning in the Search for New Fundamental Physicscited by 50
  2. 2021Online-compatible Unsupervised Non-resonant Anomaly Detectioncited by 15
  3. 2022Snowmass 2021 Computational Frontier CompF03 Topical Group Report: Machine Learningcited by 10
  4. 2022Interpretable Uncertainty Quantification in AI for HEPcited by 8
  5. 2022Bias and Priors in Machine Learning Calibrations for High Energy Physicscited by 3
  6. 2025Evidential Deep Learning for Uncertainty Quantification and Out-of-Distribution Detection in Jet Identification using Deep Neural Networkscited by 3
  7. 2026Uncertainty in Physics and AI: Taxonomy, Quantification, and Validationcited by 3
  8. 2024FAIR Universe HiggsML Uncertainty Dataset and Competitioncited by 2
  9. 2022New Metric Formulas that Include Measurement Errors in Machine Learning for Natural Sciencescited by 1
  10. 2024Enhancing generalization in high energy physics using white-box adversarial attackscited by 0
  11. 2026Functional anatomy of Pythia-Herwig differences with Kolmogorov-Arnold networkscited by 0