Regensburg 2007 – scientific programme

Parts | Days | Selection | Search | Downloads | Help

DY: Fachverband Dynamik und Statistische Physik

DY 28: Nonlinear stochastic systems

DY 28.3: Talk

Thursday, March 29, 2007, 14:30–14:45, H2

Generalized Markov approximations — •Detlef Holstein and Holger Kantz — Max-Planck-Institut fuer Physik komplexer Systeme, Dresden

In general arbitrary dynamics can be highly nonmarkovian, i.e. the future evolution of the dynamics can in principle depend on quite a lot of time points in the far past. For practical purposes it is often necessary to truncate a rather irrelevant part of the memory. Hence a criterion based on information theory and statistics is developed, which identifies the relevant memory terms. Consequences for prediction, especially wind speed prediction, are investigated.