Catálogo de publicaciones - libros

In the exponential regression model with an autoregressive error structure exact D -optimal designs for weighted least squares analysis are investigated. It is shown that support points of a locally D -optimal design are discontinuous with respect to the correlation parameter. Also equidistant designs are proved to be considerably less efficient than maximin efficient D -optimal designs. A tool used in the study is the functional approach described in a recent book Melas (2006).

The aim of the paper is to present a method of sample size determination for the Kruskal — Wallis test. The method is based on the concept of the relative effect between the two extreme distributions of those sampled and on the maxi-min size for the usual F -test.

In this article we extend the theory of optimum designs for generalized linear models, addressing the optimality of designs for parameter estimation in a location-dispersion model when either not all p parameters in the mean model or not all q parameters in the dispersion model are of interest. The criterion of Bayesian D _s-optimality is adopted and its properties are derived. The theory is illustrated with an example from the coffee industry.

The purpose of this note is to show by a simple example that some of the favourite results in optimal design theory do not necessarily carry over if random effects are involved. In particular, the usage of the popular D -criterion appears to be doubtful.

cDNA microarrays are a powerful tool in gene expression analysis Speed (2003). Landgrebe et al (2006) proposed a special 3-factor model to estimate various effects on the log ratios of measured fluorescence intensities. We demonstrate in this paper that the Within-B-Swap (BS) design introduced by Landgrebe et al (2006) is A- and D-optimal for estimating the linear contrast for the treatment effect in the general case of l treatments and k cell lines.

In this paper we discuss the structure of the information matrices of D -optimal experimental designs for the parameters in a stationary process when the parametrized correlation structure satisfies mild conditions. Such conditions are easily fulfilled by many correlation structures, e.g. structures from power exponential family and some members of the Matérn class. We provide a lower bound for information on the mean parameter and prove it to be an increasing function of distances of design points. The design points can collapse under the presence of some covariance structures and a so called nugget effect can be employed in a natural way. We also show that the information of equidistant designs (designs with equally spaced design points) on the covariance parameter is increasing with the number of design points under our conditions on correlations. If only trend parameters are of interest, the designs covering the whole design space non-uniformly are rather efficient.

Typically T-optimality is used to discriminate among several models with Normal errors. In order to discriminate between two non-Normal models, a criterion based on the Kullback-Liebler distance has been proposed, the so called KL-criterion. In this paper, a generalization of the KL-criterion is proposed to deal with discrimination among several non-Normal models. An example where three logistic regression models are compared is provided.

Three-level factorial designs can be used to perform factor screening and subsequently response surface exploration on its projections in a single stage experiment. Here we select optimal designs for this approach from 18-run and 27-run orthogonal designs. Our choices are based on two types of design criteria. Besides commonly used model estimation criteria, we also consider model discrimination criteria.