Efficient rounding of approximate designs

Abstract

?seq=1&cid=pdf-reference#references_tab_contents You may need to log in to JSTOR to access the linked references. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact

Optimal Designs for Regression Models With a Constant Coefficient of Variation

“…. s) [see Pukelsheim and Rieder (1992)]. The analogue of the Fisher information matrix for an approximate design is given by the matrix…”

Untitled

“…Regardless of whether ξ is a continuous-or rational-mass design, nλ i is to be thought of as the number of observations taken at the experimental level s i . In regression settings, we advocate here that continuous point-continuous mass designs be obtained as a general rule-of-thumb, at least as a starting point; these designs can later be rounded to practical designs using the methodology given in [Pukelsheim and Rieder 1992]. For simplicity, we refer to these designs as continuous designs for the remainder of this paper.…”

“…Table 2 contains the continuous locally Q-optimal designs with n = 3 for this model for σ 0 = 0, 0.05, 0.10, 0.15, and 0.20. As pointed out above, techniques to obtain practical designs from design measures are discussed in [Pukelsheim and Rieder 1992].…”

Optimal discrimination designs for exponential regression models

“…+ N n = N . An optimal rounding procedure is described in Pukelsheim and Rieder (1992). In the following, we will therefore restrict ourselves to the analysis and calculation of approximate designs to avoid the problems of discrete optimization.…”