Optimal Designs with Non-exchangeable Runs
نویسنده
چکیده
Popular algorithms for optimal design do not account for non-exchangeability of the basic experimental units. This paper introduces an algorithm for searching for D-optimal experimental designs in the presence of certain kinds of non-exchangeability|in particular, when there are xed covariates associated with units and/or when the units have a non-trivial covariance structure; this algorithm extends and improves on the one of Cook and Nachtsheim (1989) for nding optimal block designs. Examples are given where the in uence of the covariates/covariance on the choice of design is known or at least roughly apparent, and the algorithm proves to be quite successful in these situations.
منابع مشابه
Generalized Resolution and Minimum Aberration for Nonregular Fractional Factorial Designs
Seeking the optimal design with a given number of runs is a main problem in fractional factorial designs(FFDs). Resolution of a design is the most widely usage criterion, which is introduced by Box and Hunter(1961), used to be employed to regular FFDs. The resolution criterion is extended to non-regular FFG, called the generalized resolution criterion. This criterion is providing the idea of ge...
متن کاملOptimal blocked minimum-support designs for non-linear models
Finding optimal designs for experiments for non-linear models and dependent data is a challenging task. We show how the problem simplifies when the search is restricted to designs that are minimally supported; that is, the number of distinct runs (treatments) is equal to the number of unknown parameters, p, in the model. Under this restriction, the problem of finding a locally or pseudo-Bayesia...
متن کاملGenerating exact D-optimal designs for polynomial models
This paper compares several optimization algorithms that can be used to generate exact D-optimal designs (i.e., designs for a specified number of runs) for any polynomial model. The merits and limitations of each algorithm are demonstrated on several low-order polynomial models, with numerical results verified against analytical results. The efficiencies – with respect to estimating model param...
متن کاملSimplifying Complex Designs: Bayes Linear Experimental Design for Grouped Multivariate Exchangeable Systems
We adopt a Bayes linear approach to tackle design problems with many variables cross-classified in many ways. We investigate designs where we wish to sample individuals belonging to different groups, exploiting the powerful properties of the adjustment of second-order exchangeable vectors. The types of information we gain by sampling are identified with the orthogonal canonical directions. We s...
متن کاملComputer-Generated Minimal (and Larger) Response-Surface Designs: (I) The Sphere
Computer-generated designs in the sphere are described which have the minimal (or larger) number of runs for a full quadratic response-surface design. In the case of 3 factors, the designs have 10 through 33 runs; for 4 factors, 15 through 28 runs; for 5 factors, 21 through 33 runs; etc. Some of these designs are listed here in full; the others can be obtained from the authors. The designs were...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 1996