mODa6, 6th International Workshop on Model-Oriented Design and Analysis


Type Congress Organisation

Funding Bodies
  • Commission of the European Communities
  • GlaxoSmithKline

Duration June 25, 2001 - June 29, 2001

http://statistik.wu-wien.ac.at/moda6/moda6.html
  • Mathematical Methods in Statistics AE (Former organization)

Tags

Press 'enter' for creating the tag
  • Hackl, Peter (Former researcher)
  • Müller, Werner (Former researcher) Project Head
  • Sedlacek, Günther (Former researcher)
 

Abstract (German)

High-level Scientific Conference:<P></P>mODa workshops have been and will continue to be driving forces of the scientific progress in the field of design of experiments.<P></P>The theory for the design of experiments has different branches. A model-oriented view, where some knowledge about the form of the data-generating process is assumed, naturally leads to so-called optimum design of experiments. Here, the main construction principle for a well designed experimental plan is as follows. Asserting hypothesis-testing or parameter estimation as the aim of the experiment, the task remains to take a (possibly given) number of measurements in a way, that either the power of the test or the precision of the estimator is maximised. Additional requirements such as the unbiasedness of the statistical method, restrictions on the experimental conditions (e.g. to a certain experimental region), etc. usually enter as side-conditions of the maximisation problem. The advantage of the optimum design approach is that it allows the application of computational algorithms for the construction of tailor-made designs.<P></P>The model-oriented point of view is what holds all mODa workshops thematically together. In the past this has led to a number of significant developments in both the theory and application of DOE which have been initiated or presented at a mODa workshop. It is astounding what a rich diversity of topics emerged from this common point of view, which continues to fertilise the subject.


Abstract (English)

High-level Scientific Conference:<P></P>mODa workshops have been and will continue to be driving forces of the scientific progress in the field of design of experiments.<P></P>The theory for the design of experiments has different branches. A model-oriented view, where some knowledge about the form of the data-generating process is assumed, naturally leads to so-called optimum design of experiments. Here, the main construction principle for a well designed experimental plan is as follows. Asserting hypothesis-testing or parameter estimation as the aim of the experiment, the task remains to take a (possibly given) number of measurements in a way, that either the power of the test or the precision of the estimator is maximised. Additional requirements such as the unbiasedness of the statistical method, restrictions on the experimental conditions (e.g. to a certain experimental region), etc. usually enter as side-conditions of the maximisation problem. The advantage of the optimum design approach is that it allows the application of computational algorithms for the construction of tailor-made designs.<P></P>The model-oriented point of view is what holds all mODa workshops thematically together. In the past this has led to a number of significant developments in both the theory and application of DOE which have been initiated or presented at a mODa workshop. It is astounding what a rich diversity of topics emerged from this common point of view, which continues to fertilise the subject.

Partners

  • University of Warwick - United Kingdom

Publications

Classification

  • 1113 Mathematical statistics (Details)

Expertise

  • optimum design