event
Thursday 30 Mar 2017: Statistical Science Seminar: Model Discrepancy
Professor Michael Goldstein - Durham University
H101 15:00-16:00
Models are widely used for studying real world
phenomena. However, even the most carefully constructed model contains
a host of guesses and approximations to the real world
problem. Therefore, an essential component of any model based analysis
is the careful structuring and assessment of the irreducible
differences between the model and the physical system. We will discuss
and illustrate various principles and methods for carrying out such a
careful discrepancy analysis.