Home

About

Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network
FAQHELPSITE MAPCONTACT US


  Advanced Search  

 
The Center for Control, Dynamical Systems, and Computation University of California at Santa Barbara
 

Summary: The Center for Control, Dynamical Systems, and Computation
University of California at Santa Barbara
Winter 2009 Seminar Series
Presents
An Approximation Theory for Model Validation
Roger Ghanem
USC
Friday, January 16, 2009 3:00-4:00pm Chem 1171
Abstract:
The availability of trustworthy analytical tools is, clearly, a critical requirement for analytical certification
or qualification and for useful prognosis. The concept of model validation has been introduced with the
aim of providing a quantitative description of the proximity between model prediction and the physical
reality it purports to represent, and can thus be viewed as a building block towards tackling challenges
in analytical certification and prognosis. A challenge in model validation has been to separate the
reasons for observed discrepancies between predictions and observations, and thus pave the way for
model improvement, in case a model fails to be valid.
This presentation will describe recent successful efforts for constructing stochastic models that permit
the segregation of some key errors associated with typical model-based predictions. These include
errors stemming from calibration using limited data, errors associated with finite sample size in a
Monte Carlo run, and errors associated with spatio-temporal discretizations of governing equations. In

  

Source: Akhmedov, Azer - Department of Mathematics, University of California at Santa Barbara

 

Collections: Mathematics