Abstract:
Optimal prediction methods compensate for a lack of resolution in the
numerical solution of time-dependent differential equations through
the use of prior statistical information. I will present a simple derivation
of the basic methodology, emphasizing nonlinear aspects. I will show that
perturbation theory provides a useful device for
dealing with quasi-linear problems, and provide nonlinear examples
that illuminates the difference between a pseudo-spectral method and an
optimal prediction method with Fourier kernels. Along the way, I will
explain the differences and similarities between optimal prediction,
data acquisition, and duality methods for
finding weak solutions.
(Joint work with A. Kast, R. Kupferman and D. Levy).
Zeit: | Dienstag, 08. Juni 1999, 17.00 Uhr |
Ort: | Konrad-Zuse-Zentrum, Takustr. 7, 14195 Berlin, Seminarraum |