Remarks on uncertainty assessment and management in modeling and computation (Q5931748)

This paper discusses some questions related to uncertainty in computations for mathematical models. A computationally tractable probabilistic framework to treat uncertainty in the estimation of parameters or inverse problem is given. An example illustrates the mathematical development. It consists to estimate constant parameters in differential equations by treating the parameters as random variables. This approach allows to attach a quantitative measure of variability to estimated parameters by viewing them as random variables.