Machine Learning

Demonstration of the relationship between sensitivity and identifiability for inverse uncertainty quantification; a comprehensive modular Bayesian approach is used for inverse uncertainty quantification. A comprehensive modular Bayesian approach is used for inverse Uncertainty Quantification. Input identifiability is closely related to its sensitivity to the outputs. “Fake identifiability” is possible if responses are not appropriately chosen. Inaccurate but informative prior distributions can also lead to “fake identifiability”.