S3RI Seminar - A design criterion for symmetric model discrimination based on nominal confidence sets, Werner Mueller (JKU Linz) Seminar

Time:: 14:00 - 15:00
Date:: 13 February 2018
Venue:: Building 44, Room 1041, University of Southampton, Highfield Campus, Southampton, SO17 1BJ

For more information regarding this seminar, please email Dr Helen Ogden at H.E.Ogden@southampton.ac.uk .

Event details

Experimental design applications for discriminating between models have been hampered by the assumption to know beforehand which model is the true one, which is counter to the very aim of the experiment. Previous approaches to alleviate this requirement were either symmetrizations of asymmetric techniques, or Bayesian, minimax and sequential approaches.

Here we present a genuinely symmetric criterion based on a linearized distance between mean-value surfaces and the newly introduced notion of nominal confidence sets. The computation of the proposed criterion leads to a specific convex optimization problem which results in a substantial increase of computational speed compared with the standard approaches. This permits direct computation of optimal exact designs, instead of the usual approximate designs. Moreover, the nominal confidence sets add flexibility to the construction of discrimination designs, allowing the experimenters to incorporate their degree of certainty about nominal values of the model parameters.

We provide a Monte-Carlo evaluation of the methods discrimination performance on the basis of the likelihood ratio. An application for a pair of competing models in enzyme kinetics is given as well.

The seminar will also be available via a live web-cast at
https://cours ecast.soton.ac. uk/Panopto/Page s/Viewer.aspx?i d=05b6e55c-968e -459b-b8c8-eaab ebc274a7

Speaker information

Werner Mueller, JKU Linz. Johannes Keplier Universitat (JKU) Linz . Werner G. Müller is currently a full professor at the Department of Applied Statistics at the Johannes Kepler University Linz, Austria (JKU) and President of the Austrian Statistical Society (ÖSG) . He gained his PhD in Statistics from the University of Vienna in 1991 and has worked as an Associate Professor at the University of Vienna until 2006. His research interests are in the areas of experimental design and spatial statistics. Author of the book Collecting Spatial Data Müller has co-edited six other books, has published more than 50 research papers in refereed journals, and has Contributed with more than 70 other publications, with currently over 2,000 citations on Google Scholar. He serves as Coeditor of Statistical Papers (Springer) since 2010.