Designs for generalized linear models with several variables and model uncertainty

Standard factorial designs sometimes may be inadequate for experiments that aim to estimate a generalized linear model, for example, for describing a binary response in terms of several variables. A method is proposed for finding exact designs for such experiments that uses a criterion allowing for uncertainty in the link function, the linear predictor, or the model parameters, together with a design search. Designs are assessed and compared by simulation of the distribution of efficiencies relative to locally optimal designs over a space of possible models. Exact designs are investigated for two applications, and their advantages over factorial and central composite designs are demonstrated.

Bridging the gap between different statistical approaches: An integrated framework for modelling

This paper proposes a template for modelling complex datasets that integrates traditional statistical modelling approaches with more recent advances in statistics and modelling through an exploratory framework. Our approach builds on the well-known and long standing traditional idea of 'good practice in statistics' by establishing a comprehensive framework for modelling that focuses on exploration, prediction, interpretation and reliability assessment, a relatively new idea that allows individual assessment of predictions. The integrated framework we present comprises two stages. The first involves the use of exploratory methods to help visually understand the data and identify a parsimonious set of explanatory variables. The second encompasses a two step modelling process, where the use of non-parametric methods such as decision trees and generalized additive models are promoted to identify important variables and their modelling relationship with the response before a final predictive model is considered. We focus on fitting the predictive model using parametric, non-parametric and Bayesian approaches. This paper is motivated by a medical problem where interest focuses on developing a risk stratification system for morbidity of 1,710 cardiac patients given a suite of demographic, clinical and preoperative variables. Although the methods we use are applied specifically to this case study, these methods can be applied across any field, irrespective of the type of response.

Application of a generalized thermodynamic approach to characterize mesoporous materials

Equilibrium adsorption and desorption in mesoporous adsorbents is considered on the basis of rigorous thermodynamic analysis, in which the curvature-dependent solid-fluid potential and the compressibility of the adsorbed phase are accounted for. The compressibility of the adsorbed phase is considered for the first time in the literature in the framework of a rigorous thermodynamic approach. Our model is a further development of continuum thermodynamic approaches proposed by Derjaguin and Broekhoff and de Boer, and it is based on a reference isotherm of a non-porous material having the same chemical structure as that of the pore wall. In this improved thermodynamic model, we incorporated a prescription for transforming the solid-fluid potential exerted by the flat reference surface to the potential inside cylindrical and spherical pores. We relax the assumption that the adsorbed film density is constant and equal to that of the saturated liquid. Instead, the density of the adsorbed fluid is allowed to vary over the adsorbed film thickness and is calculated by an equation of state. As a result, the model is capable to describe the adsorption-desorption reversibility in cylindrical pores having diameter less than 2 nm. The generalized thermodynamic model may be applied to the pore size characterization of mesoporous materials instead of much more time-consuming molecular approaches. (c) 2005 Elsevier B.V. All rights reserved.

Orbital currents and charge density waves in a generalized Hubbard ladder

We study a generalized Hubbard model on the two-leg ladder at zero temperature, focusing on a parameter region with staggered flux (SF)/d-density wave (DDW) order. To guide our numerical calculations, we first investigate the location of a SF/DDW phase in the phase diagram of the half-filled weakly interacting ladder using a perturbative renormalization group (RG) and bosonization approach. For hole doping 6 away from half-filling, finite-system density-matrix renormalizationgroup (DMRG) calculations are used to study ladders with up to 200 rungs for intermediate-strength interactions. In the doped SF/DDW phase, the staggered rung current and the rung electron density both show periodic spatial oscillations, with characteristic wavelengths 2/delta and 1/delta, respectively, corresponding to ordering wavevectors 2k(F) and 4k(F) for the currents and densities, where 2k(F) = pi(1 - delta). The density minima are located at the anti-phase domain walls of the staggered current. For sufficiently large dopings, SF/DDW order is suppressed. The rung density modulation also exists in neighboring phases where currents decay exponentially. We show that most of the DMRG results can be qualitatively understood from weak-coupling RG/bosonization arguments. However, while these arguments seem to suggest a crossover from non-decaying correlations to power-law decay at a length scale of order 1/delta, the DMRG results are consistent with a true long-range order scenario for the currents and densities. (c) 2005 Elsevier Inc. All rights reserved.

Simulating the effects of dominance and epistasis on selection response in the CIMMYT wheat breeding program using QuCim

Plant breeders use many different breeding methods to develop superior cultivars. However, it is difficult, cumbersome, and expensive to evaluate the performance of a breeding method or to compare the efficiencies of different breeding methods within an ongoing breeding program. To facilitate comparisons, we developed a QU-GENE module called QuCim that can simulate a large number of breeding strategies for self-pollinated species. The wheat breeding strategy Selected Bulk used by CIMMYT's wheat breeding program was defined in QuCim as an example of how this is done. This selection method was simulated in QuCim to investigate the effects of deviations from the additive genetic model, in the form of dominance and epistasis, on selection outcomes. The simulation results indicate that the partial dominance model does not greatly influence genetic advance compared with the pure additive model. Genetic advance in genetic systems with overdominance and epistasis are slower than when gene effects are purely additive or partially dominant. The additive gene effect is an appropriate indicator of the change in gene frequency following selection when epistasis is absent. In the absence of epistasis, the additive variance decreases rapidly with selection. However, after several cycles of selection it remains relatively fixed when epistasis is present. The variance from partial dominance is relatively small and therefore hard to detect by the covariance among half sibs and the covariance among full sibs. The dominance variance from the overdominance model can be identified successfully, but it does not change significantly, which confirms that overdominance cannot be utilized by an inbred breeding program. QuCim is an effective tool to compare selection strategies and to validate some theories in quantitative genetics.

Comparison of virulence gene profiles of Escherichia coli strains isolated from healthy and diarrheic swine

A combination of uni- and multiplex PCR assays targeting 58 virulence genes (VGs) associated with Escherichia coli strains causing intestinal and extraintestinal disease in humans and other mammals was used to analyze the VG repertoire of 23 commensal E. coli isolates from healthy pigs and 52 clinical isolates associated with porcine neonatal diarrhea (ND) and postweaning diarrhea (PWD). The relationship between the presence and absence of VGs was interrogated using three statistical methods. According to the generalized linear model, 17 of 58 VGs were found to be significant (P < 0.05) in distinguishing between commensal and clinical isolates. Nine of the 17 genes represented by iha, hlyA, aidA, east1, aah, fimH, iroN(E).(coli), traT, and saa have not been previously identified as important VGs in clinical porcine isolates in Australia. The remaining eight VGs code for fimbriae (F4, F5, F18, and F41) and toxins (STa, STh, LT, and Stx2), normally associated with porcine enterotoxigenic E. coli. Agglomerative hierarchical algorithm analysis grouped E. coli strains into subclusters based primarily on their serogroup. Multivariate analyses of clonal relationships based on the 17 VGs were collapsed into two-dimensional space by principal coordinate analysis. PWD clones were distributed in two quadrants, separated from ND and commensal clones, which tended to cluster within one quadrant. Clonal subclusters within quadrants were highly correlated with serogroups. These methods of analysis provide different perspectives in our attempts to understand how commensal and clinical porcine enterotoxigenic E. coli strains have evolved and are engaged in the dynamic process of losing or acquiring VGs within the pig population.

Specific heat capacity of Australian honeys from 35 to 165C as a function of composition using differential scanning calorimetry

Modulated temperature differential scanning calorimetry was used to investigate the specific heat capacity (C-p) of 10 Australian honeys within the processing and handling temperatures. The values obtained were found to be different from the literature values at certain temperatures, and are not predictable by the additive model. The C-p of each honey exhibited a cubic relationship (P < 0.001) with the temperature (T, C). In addition, the moisture (M, %), fructose (F, %) and glucose (G, %) contents of the honeys influenced their C-p. The following equation (r(2) = 0.92) was proposed for estimating C-p of honey, and is recommended for use in the honey industry and in research: C = 996.7 + 1.4 x 10(-3)T + 5.6 x 10(-5)T(2) - 2.4 x 10(-7)T(3) - 56.5M - 25.8F - 31.0G + 1.5(M * F) + 1.8(M * G) + 0.8(F * G) - 4.6 x 10(-2) (M * F * G).

Observation of generalized synchronization of chaos in a driven chaotic system

We report on the experimental observation of the generalized synchronization of chaos in a real physical system. We show that under a nonlinear resonant interaction, the chaotic dynamics of a single mode laser can become functionally related to that of a chaotic driving signal and furthermore as the coupling strength is further increased, the chaotic dynamics of the laser approaches that of the driving signal.

Production under uncertainty and choice under uncertainty in the emergence of generalized expected utility theory

This paper presents a personal view of the interaction between the analysis of choice under uncertainty and the analysis of production under uncertainty. Interest in the foundations of the theory of choice under uncertainty was stimulated by applications of expected utility theory such as the Sandmo model of production under uncertainty. This interest led to the development of generalized models including rank-dependent expected utility theory. In turn, the development of generalized expected utility models raised the question of whether such models could be used in the analysis of applied problems such as those involving production under uncertainty. Finally, the revival of the state-contingent approach led to the recognition of a fundamental duality between choice problems and production problems.

Characterization of pore wall heterogeneity in nanoporous carbons using adsorption: the slit pore model revisited

In this paper, we propose a new nonlocal density functional theory characterization procedure, the finite wall thickness model, for nanoporous carbons, whereby heterogeneity of pore size and pore walls in the carbon is probed simultaneously. We determine the pore size distributions and pore wall thickness distributions of several commercial activated carbons and coal chars, with good correspondence with X-ray diffraction. It is shown that the conventional infinite wall thickness approach overestimates the pore size slightly. Pore-pore correlation has been shown to have a negligible effect on prediction of pore size and pore wall thickness distributions for small molecules such as argon used in characterization. By utilizing the structural parameters (pore size and pore wall thickness distribution) in the generalized adsorption isotherm (GAI) we are able to predict adsorption uptake of supercritical gases in BPL and Norit RI Extra carbons, in excellent agreement with experimental adsorption uptake data up to 60 MPa. The method offers a useful technique for probing features of the solid skeleton, hitherto studied by crystallographic methods.

Estimating Income Inequality in China Using Grouped Data and the Generalized Beta Distribution

There are two main types of data sources of income distributions in China: household survey data and grouped data. Household survey data are typically available for isolated years and individual provinces. In comparison, aggregate or grouped data are typically available more frequently and usually have national coverage. In principle, grouped data allow investigation of the change of inequality over longer, continuous periods of time, and the identification of patterns of inequality across broader regions. Nevertheless, a major limitation of grouped data is that only mean (average) income and income shares of quintile or decile groups of the population are reported. Directly using grouped data reported in this format is equivalent to assuming that all individuals in a quintile or decile group have the same income. This potentially distorts the estimate of inequality within each region. The aim of this paper is to apply an improved econometric method designed to use grouped data to study income inequality in China. A generalized beta distribution is employed to model income inequality in China at various levels and periods of time. The generalized beta distribution is more general and flexible than the lognormal distribution that has been used in past research, and also relaxes the assumption of a uniform distribution of income within quintile and decile groups of populations. The paper studies the nature and extent of inequality in rural and urban China over the period 1978 to 2002. Income inequality in the whole of China is then modeled using a mixture of province-specific distributions. The estimated results are used to study the trends in national inequality, and to discuss the empirical findings in the light of economic reforms, regional policies, and globalization of the Chinese economy.

Long-term survivor mixture model with random effects: application to a multi-centre clinical trial of carcinoma

A mixture model incorporating long-term survivors has been adopted in the field of biostatistics where some individuals may never experience the failure event under study. The surviving fractions may be considered as cured. In most applications, the survival times are assumed to be independent. However, when the survival data are obtained from a multi-centre clinical trial, it is conceived that the environ mental conditions and facilities shared within clinic affects the proportion cured as well as the failure risk for the uncured individuals. It necessitates a long-term survivor mixture model with random effects. In this paper, the long-term survivor mixture model is extended for the analysis of multivariate failure time data using the generalized linear mixed model (GLMM) approach. The proposed model is applied to analyse a numerical data set from a multi-centre clinical trial of carcinoma as an illustration. Some simulation experiments are performed to assess the applicability of the model based on the average biases of the estimates formed. Copyright (C) 2001 John Wiley & Sons, Ltd.

Three-dimensional orthotropic viscoelastic finite element model of a human ligament

Ligaments undergo finite strain displaying hyperelastic behaviour as the initially tangled fibrils present straighten out, combined with viscoelastic behaviour (strain rate sensitivity). In the present study the anterior cruciate ligament of the human knee joint is modelled in three dimensions to gain an understanding of the stress distribution over the ligament due to motion imposed on the ends, determined from experimental studies. A three dimensional, finite strain material model of ligaments has recently been proposed by Pioletti in Ref. [2]. It is attractive as it separates out elastic stress from that due to the present strain rate and that due to the past history of deformation. However, it treats the ligament as isotropic and incompressible. While the second assumption is reasonable, the first is clearly untrue. In the present study an alternative model of the elastic behaviour due to Bonet and Burton (Ref. [4]) is generalized. Bonet and Burton consider finite strain with constant modulii for the fibres and for the matrix of a transversely isotropic composite. In the present work, the fibre modulus is first made to increase exponentially from zero with an invariant that provides a measure of the stretch in the fibre direction. At 12% strain in the fibre direction, a new reference state is then adopted, after which the material modulus is made constant, as in Bonet and Burton's model. The strain rate dependence can be added, either using Pioletti's isotropic approximation, or by making the effect depend on the strain rate in the fibre direction only. A solid model of a ligament is constructed, based on experimentally measured sections, and the deformation predicted using explicit integration in time. This approach simplifies the coding of the material model, but has a limitation due to the detrimental effect on stability of integration of the substantial damping implied by the nonlinear dependence of stress on strain rate. At present, an artificially high density is being used to provide stability, while the dynamics are being removed from the solution using artificial viscosity. The result is a quasi-static solution incorporating the effect of strain rate. Alternate approaches to material modelling and integration are discussed, that may result in a better model.

A zero-augmented gamma mixed model for longitudinal data with many zeros

In many occupational safety interventions, the objective is to reduce the injury incidence as well as the mean claims cost once injury has occurred. The claims cost data within a period typically contain a large proportion of zero observations (no claim). The distribution thus comprises a point mass at 0 mixed with a non-degenerate parametric component. Essentially, the likelihood function can be factorized into two orthogonal components. These two components relate respectively to the effect of covariates on the incidence of claims and the magnitude of claims, given that claims are made. Furthermore, the longitudinal nature of the intervention inherently imposes some correlation among the observations. This paper introduces a zero-augmented gamma random effects model for analysing longitudinal data with many zeros. Adopting the generalized linear mixed model (GLMM) approach reduces the original problem to the fitting of two independent GLMMs. The method is applied to evaluate the effectiveness of a workplace risk assessment teams program, trialled within the cleaning services of a Western Australian public hospital.