The parameters uncertainty inflation fallacy

Abstract

Statistical estimation of the prediction uncertainty of physical models is typically hindered by the inadequacy of these models due to various approximations they are built upon. The prediction errors due to model inadequacy can be handled either by correcting the model's results, or by adapting the model's parameters uncertainty to generate prediction uncertainty representative, in a way to be defined, of model inadequacy errors. The main advantage of the latter approach is its transferability to the prediction of other quantities of interest based on the same parameters. A critical review of state-of-the-art implementations of this approach in computational chemistry shows that it is biased, in the sense that it does not produce prediction uncertainty bands conforming with model inadequacy errors.

Post-print version

References in corpus (4)

• Systematic Error Estimation for Chemical Reaction Energies
• Validating Predictions of Unobserved Quantities
• A critical review of statistical calibration/prediction models handling data inconsistency and model inadequacy
• Reliable estimation of prediction uncertainty for physico-chemical property models

Cited by in corpus (22)

• The Exploration of Chemical Reaction Networks
• Uncertainty Quantification in Classical Molecular Dynamics
• Bayesian machine learning for quantum molecular dynamics
• Error-Controlled Exploration of Chemical Reaction Networks with Gaussian Processes
• A critical review of statistical calibration/prediction models handling data inconsistency and model inadequacy
• Reliable estimation of prediction uncertainty for physico-chemical property models
• Universal QM/MM Approaches for General Nanoscale Applications
• Mechanism Deduction from Noisy Chemical Reaction Networks
• Error Assessment of Computational Models in Chemistry
• Probabilistic performance estimators for computational chemistry methods: the empirical cumulative distribution function of absolute errors
• The long road to calibrated prediction uncertainty in computational chemistry
• Molecule-specific Uncertainty Quantification in Quantum Chemical Studies
• Statistical Analysis of Semiclassical Dispersion Corrections
• Probabilistic performance estimators for computational chemistry methods: Systematic Improvement Probability and Ranking Probability Matrix. I. Theory
• Prediction uncertainty validation for computational chemists
• Uncertainty quantification for classical effective potentials: an extension to potfit
• Bayesian optimization for inverse problems in time-dependent quantum dynamics
• Heuristics and Uncertainty Quantification in Rational and Inverse Compound and Catalyst Design
• Stochastic Subspace via Probabilistic Principal Component Analysis for Characterizing Model Error
• Embedded Model Form Uncertainty Quantification with Measurement Noise for Bayesian Model Calibration
• Extending OpenKIM with an Uncertainty Quantification Toolkit for Molecular Modeling
• Inverse design of bespoke interatomic potentials via active learning by information-matching