Applied and Computational Math Seminar: Estimating parameters of Ito diffusion using the fluctuation dissipation theory response statistics

Speaker:John Harlim, Penn State University
Title: Estimating parameters of Ito diffusion using the fluctuation dissipation theory response statistics

Abstract: I will discuss a new parameter estimation method for Ito diffusions such that the resulting model predicts the equilibrium statistics as well as the sensitivities of the underlying system to external disturbances. Our formulation does not require the knowledge of the underlying system, however, we assume that the linear response statistics can be computed via the fluctuation-dissipation theory. The main idea is to fit the model to a finite set of ``essential" statistics that is sufficient to approximate the linear response operators. I will discuss the consistency of the method and the numerical performance of the scheme applied on a class of Langevin equations and stochastic gradient flows.

Time: Friday, November 10, 2017, 1:30-2:30pm

Place: Exploratory Hall, Room 4106