The present book is devoted to studying optimal experimental designs for a wide class of linear and nonlinear regression models. This class includes polynomial, trigonometrical, rational, and exponential models as well as many particular models used in ecology and microbiology. As the criteria of optimality, the well known D-, E-, and c-criteria are implemented. The main idea of the book is to study the dependence of optimal - signs on values of unknown parameters and on the bounds of the design interval. Such a study can be performed on the base of the Implicit Fu- tion Theorem, the classical result of functional analysis. The idea was ?rst introduced in the author's paper (Melas, 1978) for nonlinear in parameters exponential models. Recently, it was developed for other models in a n- ber of works (Melas (1995, 2000, 2001, 2004, 2005), Dette, Melas (2002, 2003), Dette, Melas, Pepelyshev (2002, 2003, 2004b), and Dette, Melas, Biederman (2002)). Thepurposeofthepresentbookistobringtogethertheresultsobtained and to develop further underlying concepts and tools. The approach, m- tioned above, will be called the functional approach. Its brief description can be found in the Introduction. The book contains eight chapters. The ?rst chapter introduces basic concepts and results of optimal design theory, initiated mainly by J.Kiefer.
Publisher: Springer-Verlag New York Inc.
Number of pages: 338
Weight: 1010 g
Dimensions: 210 x 148 x 19 mm
Edition: 2006 ed.
From the reviews:
"This monograph is a welcome addition to the statistical literature on optimal experimental designs. ... This will be a useful reference book for researchers in this area." (Damaraju Raghavarao, Mathematical Reviews, Issue 2006 k)
"The material presented is at a sophisticated mathematical level, with a strong emphasis on the construction to why these strategies create designs that are desirable for practical implementation. The unified approach helps consolidate the individual pieces that the author has published previously, and provides greater insights into the applicability of the methods." (Christine M. Anderson-Cook, Journal of the American Statistical Association, Vol. 102, No. 477, 2007)