Robust optimal designs to a misspecified model
Chiara Tommasi, University of Milano
Download the Paper (PDF format) - September 23, 2008
Tell a colleague about it.
Printing Tips: Select 'print as image' in the Acrobat print dialog if you have trouble printing.
ABSTRACT:
Usually, in the Theory of Optimal Experimental Design the model is assumed to be known at the design stage.
In practice, however, more competing models may be plausible for the same data.
Thus, a possibility is to find an optimal design which take both model discrimination and
parameter estimation into consideration.
In this paper we follow a different approach: we find a design which is optimum for estimation purposes
but is also robust to a misspecified model. In other words, the optimum design is "good" for estimating
the unknown parameters even if the assumed model is not correct.
SUGGESTED CITATION:
Chiara Tommasi,
"Robust optimal designs to a misspecified model"
(September 2008).
UNIMI - Research Papers in Economics, Business, and Statistics.
Statistics and Mathematics.
Working Paper 37.
http://services.bepress.com/unimi/statistics/art37
Paper presented by S.M. Iacus.
| 
