Optimal designs for minimising covariances among parameter estimators in a linear model

We construct approximate optimal designs for minimising absolute covariances between least‐squares estimators of the parameters (or linear functions of the parameters) of a linear model, thereby rendering relevant parameter estimators approximately uncorrelated with each other. In particular, we consider first the case of the covariance between two linear combinations. We also consider the case of two such covariances. For this we first set up a compound optimisation problem which we transform to one of maximising two functions of the design weights simultaneously. The approaches are formulated for a general regression model and are explored through some examples including one practical problem arising in chemistry.     

A stochastic representation and sampling algorithm for nested Archimedean copulas

A general sampling algorithm for nested Archimedean copulas was recently suggested. It is given in two different forms, a recursive or an explicit one. The explicit form allows for a simpler version of the algorithm which is numerically more stable and faster since less function evaluations are required. The algorithm can also be given in general form, not being restricted to a particular nesting such as fully nested Archimedean copulas. Further, several examples are given.     

Predicting online gambling self-exclusion: an analysis of the performance of supervised machine learning models

As gambling operators become increasingly sophisticated in their analysis of individual gambling behaviour, this study evaluates the potential for using machine learning techniques to identify individuals who used self-exclusion tools out of a sample of 845 online gamblers, based on analysing trends in their gambling behaviour. Being able to identify other gamblers whose behaviour is similar to those who decided to use self-exclusion tools could, for instance, be used to share responsible gaming messages or other information that aids self-aware gambling and reduces the risk of adverse outcomes. However, operators need to understand how accurate models can be and which techniques work well. The purpose of the article is to identify the most accurate technique out of four highly diverse techniques and to discuss how to deal analytically and practically with a rare event like self-exclusion, which was used by fewer than 1% of gamblers in our data-set. We conclude that balanced training data-sets are necessary for creating effective models and that, on our data-set, the most effective method is the random forest technique which achieves an accuracy improvement of 35 percentage points versus baseline estimates.     

Macroeconomic Forecasting Using Pooled International Data

The genetic algorithm is examined as a method for solving optimization problems in econometric estimation. It does not restrict either the form or regularity of the objective function, allows a reasonably large parameter space, and does not rely on a point-to-point search. The performance is evaluated through two sets of experiments on standard test problems as well as econometric problems from the literature. First, alternative genetic algorithms that vary over mutation and crossover rates, population sizes, and other features are contrasted. Second, the genetic algorithm is compared to Nelder–Mead simplex, simulated annealing, adaptive random search, and MSCORE.     

Differential evolutionary algorithm in the construction process of optimal experimental designs

This article shows how a differential evolution algorithm can be used to find statistical designs under several optimality criteria as A, D, and T. The general algorithm of differential evolution is described and then applied on constructing optimal designs for several well-known models and compare them with those already available by other algorithms, in terms of relative efficiency. Moreover, the great effect of population size in the running of this algorithm establishes a precedent in the use of differential evolution algorithms over genetic algorithms.     

Use of Quality-Adjusted Life Year Weights with Dose-Response Models for Public Health Decisions: A Case Study of the Risks and Benefits of Fish Consumption

Risks associated with toxicants in food are often controlled by exposure reduction. When exposure recommendations are developed for foods with both harmful and beneficial qualities, however, they must balance the associated risks and benefits to maximize public health. Although quantitative methods are commonly used to evaluate health risks, such methods have not been generally applied to evaluating the health benefits associated with environmental exposures. A quantitative method for risk-benefit analysis is presented that allows for consideration of diverse health endpoints that differ in their impact (i.e., duration and severity) using dose-response modeling weighted by quality-adjusted life years saved. To demonstrate the usefulness of this method, the risks and benefits of fish consumption are evaluated using a single health risk and health benefit endpoint. Benefits are defined as the decrease in myocardial infarction mortality resulting from fish consumption, and risks are defined as the increase in neurodevelopmental delay (i.e., talking) resulting from prenatal methylmercury exposure. Fish consumption rates are based on information from Washington State. Using the proposed framework, the net health impact of eating fish is estimated in either a whole population or a population consisting of women of childbearing age and their children. It is demonstrated that across a range of fish methylmercury concentrations (0-1 ppm) and intake levels (0-25 g/day), individuals would have to weight the neurodevelopmental effects 6 times more (in the whole population) or 250 times less (among women of child-bearing age and their children) than the myocardial infarction benefits in order to be ambivalent about whether or not to consume fish. These methods can be generalized to evaluate the merits of other public health and risk management programs that involve trade-offs between risks and benefits.     

Recursive formulae for the average efficiency factor in block and row-column designs

Interchange algorithms are widely used to construct efficient block and row-column designs. We provide simple recursive formulae for updating this average efficiency factor, so that it is no longer computationally expensive to calculate it after each possible interchange.     

Ethnic Differences in Risk from Mercury among Savannah River Fishermen

Fishing plays an important role in people's lives and contaminant levels in fish are a public health concern. Many states have issued consumption advisories; South Carolina and Georgia have issued them for the Savannah River based on mercury and radionuclide levels. This study examined ethnic differences in risk from mercury exposure among people consuming fish from the Savannah River, based on site-specific consumption patterns and analysis of mercury in fish. Among fish, there were significant interspecies differences in mercury levels, and there were ethnic differences in consumption patterns. Two methods of examining risk are presented: (1) Hazard Index (HI), and (2) estimates of how much and how often people of different body mass can consume different species of fish. Blacks consumed more fish and had higher HIs than Whites. Even at the median consumption, the HI for Blacks exceeded 1.0 for bass and bowfin, and, at the 75th percentile of consumption, the HI exceeded 1.0 for almost all species. At the White male median consumption, noHI exceeded 1, but for the 95th percentile consumer, the HI exceeded 1.0 almost regardless of which species were eaten. Although females consumed about two thirds the quantity of males, HIs exceeded 1 for most Black females and for White females at or above the 75th percentile of consumption. Thus, close to half of the Black fishermen were eating enough Savannah River fish to exceed HI = 1. Caution must be used in evaluating an HI because the RfDs were developed to protect the most vulnerable individuals. The percentage of each fish species tested that exceeded the maximum permitted limits of mercury in fish was also examined. Over 80% of bowfin, 38% of bass, and 21% of pickerel sampled exceeded 0.5 ppm. The risk methodology is applicable anywhere that comparable data can be obtained. The risk estimates are representative for fishermen along the Savannah River, and are not necessarily for the general populations.     

On-line expectation–maximization algorithm for latent data models

Summary.  We propose a generic on-line (also sometimes called adaptive or recursive) version of the expectation–maximization (EM) algorithm applicable to latent variable models of independent observations. Compared with the algorithm of Titterington, this approach is more directly connected to the usual EM algorithm and does not rely on integration with respect to the complete-data distribution. The resulting algorithm is usually simpler and is shown to achieve convergence to the stationary points of the Kullback–Leibler divergence between the marginal distribution of the observation and the model distribution at the optimal rate, i.e. that of the maximum likelihood estimator. In addition, the approach proposed is also suitable for conditional (or regression) models, as illustrated in the case of the mixture of linear regressions model.     

Manufacturing cell formation using similarity coefficients and pair-wise interchange: formulation and comparison

Many algorithms have been proposed to form manufacturing cells from component routeings. Most of these methods require specialized algorithms for implementation. Some others use well-known procedures such as integer programming. But these may be difficult for practising managers to comprehend. In this study we propose a simple method that can be implemented using spreadsheet software and an inexpensive layout package such as CRAFT. In addition, we also compare our procedure with many existing procedures using eight well-known problems from the literature. The results show that the proposed procedure compares well with the existing procedures using three evaluation measures. Therefore, this procedure may be useful to practitioners and researchers.