Stochastic model reduction applied to inverse problems
نویسندگان
چکیده
This article describes the use of Gaussian Processes in model reduction techniques with application to inverse problems. Mainly, the work is focused on the proper construction of the model approximation, namely on training process based on minimal number of learning samples, by making use of automatic samples selection through computed standard deviation of model prediction. An example of application of stochastic surrogate model for the paperboard characterization through biaxial tensile test and DIC measurements is also presented.
منابع مشابه
A Stochastic Method for Solving Inverse Problems in Epidemic Modelling
We describe a stochastic optimization method that can be used to solve inverse problems in epidemic modelling. Although in general it cannot be expected that these inverse problems have solutions and that possible solutions can be approximated by local optimization methods, our optimization method worked well when applied to a model for the epidemics caused by Respiratory Syncytial Virus (RSV) ...
متن کاملAn efficient Bayesian inference approach to inverse problems based on adaptive sparse grid collocation method
A new approach for modeling inverse problems using a Bayesian inference method is introduced. The Bayesian approach considers the unknown parameters as random variables and seeks the probabilistic distribution of the unknowns. By introducing the concept of the stochastic prior state space to the Bayesian formulation, we reformulate the deterministic forward problem as a stochastic one. The adap...
متن کاملIdentification - - Inverse Problems for Partial Differential Equations : a Stochastic Formulation
This paper presents a stochastic formulation of a class of identification problems for partial differential equations, known as 'inverse' problems in the mathematicalphysics literature. By introducing stochastic processes to model errors in observation as well as 'disturbance' we can provide a precise formulation to interpret what appear to be 'ad hoc' techniques, especially in the treatment of...
متن کاملModel Order Reduction Techniques for Uncertainty Quantification Problems
The last few years have witnessed a tremendous development of the computational field of uncertainty quantification (UQ), which includes statistical, sensitivity and reliability analyses, stochastic or robust optimal control/design/optimization, parameter estimation, data assimilation, to name just a few. In all these problems, the solution of stochastic partial differential equations (PDEs) is...
متن کاملForecasting with imperfect models, dynamically constrained inverse problems, and gradient descent algorithms
This paper is part of a larger study investigating the meaning of, and appropriate procedures for, forecasting with imperfect models. (In the author’s opinion there is currently no satisfactory general theory and practice for doing so with complex nonlinear systems.) The focus of this paper is on initialisation of the forecast. At the heart of every forecasting scheme there is an inverse proble...
متن کامل