143
reads

Mong-Na Lo Huang

2012-03-02
13:00:00 - 14:40:00

103 , Mathematics Research Center Building (ori. New Math. Bldg.)

In this talk, theory of optimal experimental designs for regression models will be presented. Firstly, important principles and methodologies of finding optimal designs for regression models, such as optimality criteria (D, A, E and c), and the celebrated general equivalence theorem by Kiefer and Wolfowitz (1959), will be introduced. Then examples of optimal designs will be presented for linear regression models (polynomial, weighted polynomial) on different design spaces including the closed interval, cube, sphere and simplex etc. Later optimal designs for non-linear regression models will also be discussed.