Design for linear regression models with correlated errors (Q2799889)

This paper presents an excellent survey of results on experimental design for linear regression models with correlated errors. Old results by Sacks and Ylvisaker (1966, 1968), Bickel and Herzberg (1979) and Nöther (1985) are discussed as well as recently results by Pazman and Müller (2001), Fedorov and Müller (2007) and by the authors themselves: Dette, Leonenko, Pepelyshev and Zhigljavsky (2009); Zhigljavsky, Dette and Pepelyshev (2010) and Dette, Pepelyshev and Zhigljavsky (2013). The structure of the paper is as follows: Section 1 sketches the main approaches, Section 2 considers designs for one-parameter models, Section 3 discusses optimal designs for the BLUE and Section 4 for the OLS.NEWLINENEWLINEFor the entire collection see [Zbl 1327.62001].