SciCADE 2013
International Conference on Scientific Computation and Differential Equations
September 16-20, 2013, Valladolid (Spain)

Invited Talk

Show full list of Invited Contributions to Minisymposia Show talk context (MS03)


Consistent inference for coarse-grained models from multiscale data

S. Krumscheid, G.A. Pavliotis and S. Kalliadasis

Abstract
Most dynamical systems in the natural sciences are characterized by the presence of processes that occur across several length and time scales. Examples include the atmosphere-ocean system, biological systems, materials and molecular dynamics. Typically only the dynamics at the macroscopic scale is of interest. While multiscale methods (e.g. homogenization) provide the analytical framework for the rigorous derivation of effective coarse-grained dynamical systems, statistical inference for these multiscale systems (i.e. identifying parameters in the coarse-grained system from data of the macroscopic component) remains far from being straightforward. In particular, standard statistical techniques such as maximum likelihood become biased due to the multiscale error. In this talk we will introduce a novel class of estimators for multiscale diffusions that do not suffer from this bias. In addition to presenting rigorous convergence results, we will present several illustrative examples.

Organized by         Universidad de Valladolid     IMUVA