Bridge Designs for Modeling Systems With Low Noise

For deterministic computer simulations, Gaussian process models are a standard procedure for fitting data. These models can be used only when the study design avoids having replicated points. This characteristic is also desirable for one-dimensional projections of the design, since it may happen that one of the design factors has a strongly nonlinear effect on the response. Latin hypercube designs have uniform one-dimensional projections, but are not efficient for fitting low-order polynomials when there is a small error variance. D-optimal designs are very efficient for polynomial fitting but have substantial replication in projections. We propose a new class of designs that bridge the gap between D-optimal designs and D-optimal Latin hypercube designs. These designs guarantee a minimum distance between points in any one-dimensional projection allowing for the fit of either polynomial or Gaussian process models. Subject to this constraint they are D-optimal for a prespecified model.     

d-QPSO: A Quantum-Behaved Particle Swarm Technique for Finding D-Optimal Designs With Discrete and Continuous Factors and a Binary Response

Identifying optimal designs for generalized linear models with a binary response can be a challenging task, especially when there are both discrete and continuous independent factors in the model. Theoretical results rarely exist for such models, and for the handful that do, they usually come with restrictive assumptions. In this article, we propose the d-QPSO algorithm, a modified version of quantum-behaved particle swarm optimization, to find a variety of D-optimal approximate and exact designs for experiments with discrete and continuous factors and a binary response. We show that the d-QPSO algorithm can efficiently find locally D-optimal designs even for experiments with a large number of factors and robust pseudo-Bayesian designs when nominal values for the model parameters are not available. Additionally, we investigate robustness properties of the d-QPSO algorithm-generated designs to various model assumptions and provide real applications to design a bio-plastics odor removal experiment, an electronic static experiment, and a 10-factor car refueling experiment. Supplementary materials for the article are available online.     

D-optimal experimental design for production models in nonstandard experiments

Robust design is an effective Quality by Design method to reduce product variation by selecting levels of design factors. For a number of situations, a nonstandard design region with linearly limited resources is needed to conduct an experiment. In the literature, little attention has been given to the development of robust design models for the nonstandard design region with a combination of linearly limited resources and a limited number of design points. In this paper, a selection scheme of D-optimal experimental design points is proposed to generate design points using the modified exchange algorithm for the nonstandard design region while specifying linearly limited resources and the limited number of design points. The modified exchange algorithm is able to generate global design points with less time complexity than the improved Fedorov algorithm. In addition, robust design models linking a D-optimal experimental design with quality considerations are proposed in order to obtain optimum settings of design factors for the product. Comparative studies are also presented. Finally, a real-life experimental study shows that the proposed models with the desirability function and the sequential quadratic programming technique achieve greater variance reduction than the traditional counterparts.     

Designs in Three and Four Components For Mixtures Models With Inverse Terms

Scheffé's mixture models are given and the D-optimality of various designs for these models is discussed. Approximate D-optimal (measure) and D_n -optimal (exact) designs are given for use with mixture models with inverse terms in three and four components.     

Optimizing Two-Level Supersaturated Designs Using Swarm Intelligence Techniques

Supersaturated designs (SSDs) are often used to reduce the number of experimental runs in screening experiments with a large number of factors. As more factors are used in the study, the search for an optimal SSD becomes increasingly challenging because of the large number of feasible selection of factor level settings. This article tackles this discrete optimization problem via an algorithm based on swarm intelligence. Using the commonly used E(s²) criterion as an illustrative example, we propose an algorithm to find E(s²)-optimal SSDs by showing that they attain the theoretical lower bounds found in previous literature. We show that our algorithm consistently produces SSDs that are at least as efficient as those from the traditional CP exchange method in terms of computational effort, frequency of finding the E(s²)-optimal SSD, and also has good potential for finding D₃-, D₄-, and D₅-optimal SSDs. Supplementary materials for this article are available online.     

On Minimum-Point Second-Order Designs

In this note, we discuss k-factor, second order designs with minimum number of points ½(k + l)(k + 2), in particular, those which are extensions of designs that give minimum generalized variance for k = 2 and 3. The experimental region is the unit cuboid. Minimum point designs of this type are unknown for k ≥ 4, and these designs are the best found to date except for k = 4, where a better design is known. Kiefer has shown that these designs cannot be the best for k ≥ 7, via an existence result but, even here, specific better designs are not known and appear difficult to obtain. We also discuss some difficulties of using, in practice, designs that, are D-optimal (that is give minimum generalized variance when the number of points is not restricted).     

Using a genetic algorithm to generate D‐optimal designs for mixture‐process variable experiments

This article presents and develops a genetic algorithm (GA) to generate D‐efficient designs for mixture‐process variable experiments. It is assumed the levels of a process variable are controlled during the process. The GA approach searches design points from a set of possible points over a continuous region and works without having a finite user‐defined candidate set. We compare the performance of designs generated by the GA with designs generated by two exchange algorithms (DETMAX and k‐exchange) in terms of D‐efficiencies and fraction of design space (FDS) plots which are used to evaluate a design's prediction variance properties. To illustrate the methodology, examples involving three and four mixture components and one process variable are proposed for creating the optimal designs. The results show that GA designs have superior prediction variance properties in comparison with the DETMAX and k‐exchange algorithm designs when the design space is the simplex or is a highly‐constrained subspace of the simplex.     

An Algorithm for the Construction of “D-Optimal” Experimental Designs

This paper presenm the algorithm “DETMAX” whose purpose is to construct experimental designs that are “D-optimal.” These are designs for which the determinant of X'X is maximum, where X is the “matrix of independent variables” in the usual linear model y = Xβ + ε. Although the algorithm does not guarantee D-optimality, it has performed well in many cases where D-optimal designs are known. Five examples are given, illustrating the use of DETMAX to construct designs “from scratch” and to augment existing data. A FORTRAN listing is available on request.     

Separation in D‐optimal experimental designs for the logistic regression model

The D‐optimality criterion is often used in computer‐generated experimental designs when the response of interest is binary, such as when the attribute of interest can be categorized as pass or fail. The majority of methods in the generation of D‐optimal designs focus on logistic regression as the base model for relating a set of experimental factors with the binary response. Despite the advances in computational algorithms for calculating D‐optimal designs for the logistic regression model, very few have acknowledged the problem of separation, a phenomenon where the responses are perfectly separable by a hyperplane in the design space. Separation causes one or more parameters of the logistic regression model to be inestimable via maximum likelihood estimation. The objective of this paper is to investigate the tendency of computer‐generated, nonsequential D‐optimal designs to yield separation in small‐sample experimental data. Sets of local D‐optimal and Bayesian D‐optimal designs with different run (sample) sizes are generated for several “ground truth” logistic regression models. A Monte Carlo simulation methodology is then used to estimate the probability of separation for each design. Results of the simulation study confirm that separation occurs frequently in small‐sample data and that separation is more likely to occur when the ground truth model has interaction and quadratic terms. Finally, the paper illustrates that different designs with identical run sizes created from the same model can have significantly different chances of encountering separation.     

Sequentially Generated Second Order Quasi D-Optimal Designs for Experiments With Mixture and Process Variables

Second order designs for experiments with mixture and process variables are proposed. They are constructed on the basis of continuous D-optimal designs by use of a three-stage procedure for sequentially generating optimal designs. The determinants of the information matrices of the designs obtained are very near to those of continuous D-optimal designs. Tables of discrete quasi D-optimal designs for q + r ≤ 7 are given, where q is the number of mixture components and r is the number of process variables. The experimenter can choose the number of trials N within the interval k ≤ N ≤ min(2k, k + 20), where k is the number of model coefficients. An application of the proposed designs in an investigation of truck tire properties is given.     

The Coordinate-Exchange Algorithm for Constructing Exact Optimal Experimental Designs

We describe the cyclic coordinate-exchange algorithm for constructing D-optimal and linear-optimal experimental designs. The algorithm uses a variant of the Gauss-Southwell cyclic coordinate-descent algorithm within the k-exchange algorithm to achieve substantive reductions in required computing. Among its advantages are the following: Candidate sets, which grow exponentially in the number of factors, need not be explicitly constructed or enumerated. Convex design spaces (or mixed convex by discrete design spaces) are handled directly, without the need for sophisticated nonlinear programming routines or candidate-set adjustment. For design problems having 10 or more factors, the reductions in execution time are typically two or more orders of magnitude when compared to standard candidateset- based procedures such as k exchange, yet the designs produced exhibit no loss of efficiency.     

Two-Level Multifactor Designs for Detecting the Presence of Interactions

A design optimality criterion, tr (L)-optimality, is applied to the problem of designing two-level multifactor experiments to detect the presence of interactions among the controlled variables. We give rules for constructing tr (L)-optimal foldover designs and tr (L)-optimal fractional factorial designs. Some results are given on the power of these designs for testing the hypothesis that there are no two-factor interactions. Augmentation of the tr (L)-optimal designs produces designs that achieve a compromise between the criteria of D-optimality (for parameter estimation in a first-order model) and tr (L)-optimality (for detecting lack of fit). We give an example to demonstrate an application to the sensitivity analysis of a computer model.     

Another Look at First-Order Saturated Designs: The p-efficient Designs

The problem of constructing first-order saturated designs that are optimal in some sense has received a great deal of attention in the literature. Since these saturated designs are frequently used in screening situations, the focus will be on the potential projective models rather than the full model. This article discusses some practical concerns in choosing a design and presents some first-order saturated designs having two desirable properties, (near-) equal occurrence and (near-) orthogonality. These saturated designs are shown to be reasonably efficient for estimating the parameters of projective submodels and thus are called p-efficient designs. Comparisons with the efficiency of D-optimal designs are given for designs for all n from 3 to 30.     

Cost‐constrained G‐efficient Response Surface Designs for Cuboidal Regions

In many industrial experiments there are restrictions on the resource (or cost) required for performing the runs in a response surface design. This will require practitioners to choose some subset of the candidate set of experimental runs. The appropriate selection of design points under resource constraints is an important aspect of multi‐factor experimentation. A well‐planned experiment should consist of factor‐level combinations selected such that the resulting design will have desirable statistical properties but the resource constraints should not be violated or the experimental cost should be minimized. The resulting designs are referred to as cost‐efficient designs. We use a genetic algorithm for constructing cost‐constrained G‐efficient second‐order response surface designs over cuboidal regions when an experimental cost at a certain factor level is high and a resource constraint exists. Consideration of practical resource (or cost) restrictions and different cost structures will provide valuable information for planning effective and economical experiments when optimizing statistical design properties. Copyright © 2005 John Wiley & Sons, Ltd.     

Crashworthiness design optimization using successive response surface approximations

 Finite Element (FE) method is among the most powerful tools for crash analysis and simulation. Crashworthiness design of structural members requires repetitive and iterative application of FE simulation. This paper presents a crashworthiness design optimization methodology based on efficient and effective integration of optimization methods, FE simulations, and approximation methods. Optimization methods, although effective in general in solving structural design problems, loose their power in crashworthiness design. Objective and constraint functions in crashworthiness optimization problems are often non-smooth and highly non-linear in terms of design variables and follow from a computationally costly (FE) simulation. In this paper, a sequential approximate optimization method is utilized to deal with both the high computational cost and the non-smooth character. Crashworthiness optimization problem is divided into a series of simpler sub-problems, which are generated using approximations of objective and constraint functions. Approximations are constructed by using statistical model building technique, Response Surface Methodology (RSM) and a Genetic algorithm. The approximate optimization method is applied to solve crashworthiness design problems. These include a cylinder, a simplified vehicle and New Jersey concrete barrier optimization. The results demonstrate that the method is efficient and effective in solving crashworthiness design optimization problems. Received: 30 January 2002 / Accepted: 12 July 2002 Sponsorship for this research by the Federal Highway Administration of US Department of Transportation is gratefully acknowledged. Dr. Nielen Stander at Livermore Software Technology Corporation is also gratefully acknowledged for providing subroutines to create D-optimal experimental designs and the simplified vehicle model.     

Improved Split-Plot and Multistratum Designs

Many industrial experiments involve some factors whose levels are harder to set than others. The best way to deal with these is to plan the experiment carefully as a split-plot, or more generally a multistratum, design. Several different approaches for constructing split-plot type response surface designs have been proposed in the literature since 2001, which has allowed experimenters to make better use of their resources by using more efficient designs than the classical balanced ones. One of these approaches, the stratum-by-stratum strategy has been shown to produce designs that are less efficient than locally D-optimal designs. An improved stratum-by-stratum algorithm is given, which, though more computationally intensive than the old one, makes better use of the advantages of this approach, that is, it can be used for any structure and does not depend on prior estimates of the variance components. This is shown to be almost as good as the locally optimal designs in terms of their own criteria and more robust across a range of criteria. Supplementary materials for this article are available online.     

Optimal designs for some stochastic processes whose covariance is a function of the mean

This paper considers optimal experimental designs for models with correlated observations through a covariance function depending on the magnitude of the responses. This suggests the use of stochastic processes whose covariance structure is a function of the mean. Covariance functions must be positive definite. This fact is nontrivial in this context and constitutes one of the challenges of the present paper. We show that there exists a huge class of functions that, composed with the mean of the process in some way, preserves positive definiteness and can be used for the purposes of modeling and computing optimal designs in more realistic situations. We offer some examples for an easy construction of such covariances and then study the problem of locally D-optimal designs through an illustrative example as well as a real radiation retention model in the human body.     

Past Editors and Editorial Collaborators 1976

D-optimal fractions of three-level factorial designs for p factors are constructed for factorial effects models (2 ≤ p ≤ 4) and quadratic response surface models (2 ≤ p ≤ 5). These designs are generated using an exchange algorithm for maximizing |X′X| and an algorithm which produces D-optimal balanced array designs. The design properties for the DETMAX designs and the balanced array designs are tabulated. An example is given to illustrate the use of such designs.     

Optimal design of multi-response experiments using semi-definite programming

Optimal design of multi-response experiments for estimating the parameters of multi-response linear models is a challenging problem. The main drawback of the existing algorithms is that they require the solution of many optimization problems in the process of generating an optimal design that involve cumbersome manual operations. Furthermore, all the existing methods generate approximate design and no method for multi-response n-exact design has been cited in the literature. This paper presents a unified formulation for multi-response optimal design problem using Semi-Definite Programming (SDP) that can generate D-, A- and E-optimal designs. The proposed method alleviates the difficulties associated with the existing methods. It solves a one-shot optimization model whose solution selects the optimal design points among all possible points in the design space. We generate both approximate and n-exact designs for multi-response models by solving SDP models with integer variables. Another advantage of the proposed method lies in the amount of computation time taken to generate an optimal design for multi-response models. Several test problems have been solved using an existing interior-point based SDP solver. Numerical results show the potentials and efficiency of the proposed formulation as compared with those of other existing methods. The robustness of the generated designs with respect to the variance-covariance matrix is also investigated.