Learning about physical parameters: the importance of model discrepancy

By Jenný Brynjarsdóttir, Anthony OʼHagan. 2014

557 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computational Theory and Mathematics, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/c0c3e2873159ed15120b490d8807677168ec4aa2

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computational Theory and Mathematics

In Statistics and Probability

Citations in the corpus

A probabilistic graphical model foundation for enabling predictive digital twins at scale (2020, methodology): “Using this model we formulate an optimization problem to fit the point masses, m = [mservo, mpitot], so that the natural frequencies of the first two bending modes predicted by the model, ω1, and ω2, match the experimental data, ω̂2 and ω̂2, as closely as possible, while obeying the total mass const”
A philosophical basis for hydrological uncertainty (2016, methodology): “…by (i) estimating a model error distribution directly from data (e.g. Wilkinson et al. 2011, Sikorska et al. 2014) and then going on to test such models using an independent data source, or (ii) by performing both steps simultaneously (e.g. Vrugt and Sadegh 2013, Brynjarsdóttir and O’Hagan 2014).”
Quantification of model uncertainty in RANS simulations: A review (2018, background): “Even if there is no parameter uncertainty, so that we know the true values of all the inputs required to make a particular prediction of the process being modeled, the predicted value will not equal the true value of the process [108].”
Quantifying and reducing model-form uncertainties in Reynolds-averaged Navier-Stokes simulations: A data-driven, physics-informed Bayesian approach (2015, unknown): “is framework addresses MFU only in terms of the QoIs, whereas the modeling errors in RANS simulation arise specically from the modeled Reynolds stress term. Recent work of Brynjarsdottir and O’Hagan [11] emphasized the importance of incorporating prior information, but also highlighted the diculties”
Multi-Information Source Optimization (2016, background): “The bias f(`, x)− g(x) is also referred to as “model discrepancy” in the engineering and simulation literature [1, 4].”