BEGIN:VCALENDAR VERSION:2.0 PRODID:researchseminars.org CALSCALE:GREGORIAN X-WR-CALNAME:researchseminars.org BEGIN:VEVENT SUMMARY:Samuel Kou (Harvard University) DTSTART:20201228T003000Z DTEND:20201228T013000Z DTSTAMP:20260423T041335Z UID:iccm2020/27 DESCRIPTION:Title: Statistical inference of dynamic systems via constrained Gaussian proces ses\nby Samuel Kou (Harvard University) as part of ICCM 2020\n\n\nAbst ract\nParameter estimation of nonlinear dynamical system models from noisy and sparse experimental data is a vital task in many fields\; it has chal lenged the existing inference methods\, especially when there are unobserv ed system components. We propose a fast Bayesian inference method to estim ate the ODE parameters with real data from biological/physical experiments via constrained Gaussian process. Our method utilizes Gaussian processes that are explicitly conditioned on the functional manifold that describes the ODE system. Using this constrained Gaussian process under the Bayesian paradigm\, our method completely avoids the use of numerical solver and t hus achieves dramatic saving in computational time. At the same time\, our method also offers accurate inference\, including uncertainly quantificat ion. Our approach is distinct from the existing ones owing to its rigorous construction under the Bayesian framework. We demonstrate the speed and a ccuracy of the method using realistic examples\, including examples with u nobserved system components.\n LOCATION:https://researchseminars.org/talk/iccm2020/27/ END:VEVENT END:VCALENDAR