A critical review of statistical calibration/prediction models handling data inconsistency and model inadequacy

TL;DR

This paper critically reviews statistical calibration and prediction models that address data inconsistency and model inadequacy, evaluating their performance within a Bayesian framework using synthetic and real datasets.

Contribution

It provides a comprehensive comparison of standard and advanced statistical models for calibration and prediction under data and model uncertainties.

Findings

Bayesian models effectively handle data inconsistency and model inadequacy.

State-of-the-art models outperform traditional methods in calibration accuracy.

Evaluation on real datasets demonstrates practical applicability of the models.

Abstract

Inference of physical parameters from reference data is a well studied problem with many intricacies (inconsistent sets of data due to experimental systematic errors, approximate physical models...). The complexity is further increased when the inferred parameters are used to make predictions (virtual measurements) because parameters uncertainty has to be estimated in addition to parameters best value. The literature is rich in statistical models for the calibration/prediction problem, each having benefits and limitations. We review and evaluate standard and state-of-the-art statistical models in a common bayesian framework, and test them on synthetic and real datasets of temperature-dependent viscosity for the calibration of Lennard-Jones parameters of a Chapman-Enskog model.