A generalized estimating equations approach for analysis of the impact of new technology on a trawl fishery

The article describes a generalized estimating equations approach that was used to investigate the impact of technology on vessel performance in a trawl fishery during 1988-96, while accounting for spatial and temporal correlations in the catch-effort data. Robust estimation of parameters in the presence of several levels of clustering depended more on the choice of cluster definition than on the choice of correlation structure within the cluster. Models with smaller cluster sizes produced stable results, while models with larger cluster sizes, that may have had complex within-cluster correlation structures and that had within-cluster covariates, produced estimates sensitive to the correlation structure. The preferred model arising from this dataset assumed that catches from a vessel were correlated in the same years and the same areas, but independent in different years and areas. The model that assumed catches from a vessel were correlated in all years and areas, equivalent to a random effects term for vessel, produced spurious results. This was an unexpected finding that highlighted the need to adopt a systematic strategy for modelling. The article proposes a modelling strategy of selecting the best cluster definition first, and the working correlation structure (within clusters) second. The article discusses the selection and interpretation of the model in the light of background knowledge of the data and utility of the model, and the potential for this modelling approach to apply in similar statistical situations.

Effect of an interactive therapeutic robotic animal on engagement, mood states, agitation and psychotropic drug use in people with dementia: A cluster-randomised controlled trial protocol

Introduction: Apathy, agitated behaviours, loneliness and depression are common consequences of dementia. This trial aims to evaluate the effect of a robotic animal on behavioural and psychological symptoms of dementia in people with dementia living in long-term aged care. Methods and analysis: A cluster-randomised controlled trial with three treatment groups: PARO (robotic animal), Plush-Toy (non-robotic PARO) or Usual Care (Control). The nursing home sites are Australian Government approved and accredited facilities of 60 or more beds. The sites are located in South-East Queensland, Australia. A sample of 380 adults with a diagnosis of dementia, aged 60 years or older living in one of the participating facilities will be recruited. The intervention consists of three individual 15 min non-facilitated sessions with PARO or Plush- Toy per week, for a period of 10 weeks. The primary outcomes of interest are improvement in agitation, mood states and engagement. Secondary outcomes include sleep duration, step count, change in psychotropic medication use, change in treatment costs, and staff and family perceptions of PARO or Plush-Toy. Video data will be analysed using Noldus XT Pocket Observer; descriptive statistics will be used for participants’ demographics and outcome measures; cluster and individual level analyses to test all hypotheses and Generalised Linear Models for cluster level and Generalised Estimation Equations and/or Multi-level Modeling for individual level data. Ethics and dissemination: The study participants or their proxy will provide written informed consent. The Griffith University Human Research Ethics Committee has approved the study (NRS/03/14/HREC). The results of the study will provide evidence of the efficacy of a robotic animal as a psychosocial treatment for the behavioural and psychological symptoms of dementia. Findings will be presented at local and international conference meetings and published in peer-reviewed journals.

Inference on Cointegration in Vector Autoregressive Models (summary section only)

In the thesis we consider inference for cointegration in vector autoregressive (VAR) models. The thesis consists of an introduction and four papers. The first paper proposes a new test for cointegration in VAR models that is directly based on the eigenvalues of the least squares (LS) estimate of the autoregressive matrix. In the second paper we compare a small sample correction for the likelihood ratio (LR) test of cointegrating rank and the bootstrap. The simulation experiments show that the bootstrap works very well in practice and dominates the correction factor. The tests are applied to international stock prices data, and the .nite sample performance of the tests are investigated by simulating the data. The third paper studies the demand for money in Sweden 1970—2000 using the I(2) model. In the fourth paper we re-examine the evidence of cointegration between international stock prices. The paper shows that some of the previous empirical results can be explained by the small-sample bias and size distortion of Johansen’s LR tests for cointegration. In all papers we work with two data sets. The first data set is a Swedish money demand data set with observations on the money stock, the consumer price index, gross domestic product (GDP), the short-term interest rate and the long-term interest rate. The data are quarterly and the sample period is 1970(1)—2000(1). The second data set consists of month-end stock market index observations for Finland, France, Germany, Sweden, the United Kingdom and the United States from 1980(1) to 1997(2). Both data sets are typical of the sample sizes encountered in economic data, and the applications illustrate the usefulness of the models and tests discussed in the thesis.

Simultaneous Learning of Nonlinear Manifold and Dynamical Models for High-dimensional Time Series

The goal of this work is to learn a parsimonious and informative representation for high-dimensional time series. Conceptually, this comprises two distinct yet tightly coupled tasks: learning a low-dimensional manifold and modeling the dynamical process. These two tasks have a complementary relationship as the temporal constraints provide valuable neighborhood information for dimensionality reduction and conversely, the low-dimensional space allows dynamics to be learnt efficiently. Solving these two tasks simultaneously allows important information to be exchanged mutually. If nonlinear models are required to capture the rich complexity of time series, then the learning problem becomes harder as the nonlinearities in both tasks are coupled. The proposed solution approximates the nonlinear manifold and dynamics using piecewise linear models. The interactions among the linear models are captured in a graphical model. By exploiting the model structure, efficient inference and learning algorithms are obtained without oversimplifying the model of the underlying dynamical process. Evaluation of the proposed framework with competing approaches is conducted in three sets of experiments: dimensionality reduction and reconstruction using synthetic time series, video synthesis using a dynamic texture database, and human motion synthesis, classification and tracking on a benchmark data set. In all experiments, the proposed approach provides superior performance.

Fast and frugal models of clinical judgement in novice and expert physicians

Our objective was to study whether “compensatory” models provide better descriptions of clinical judgment than fast and frugal models, according to expertise and experience. Fifty practitioners appraised 60 vignettes describing a child with an exacerbation of asthma and rated their propensities to admit the child. Linear logistic (LL) models of their judgments were compared with a matching heuristic (MH) model that searched available cues in order of importance for a critical value indicating an admission decision. There was a small difference between the 2 models in the proportion of patients allocated correctly (admit or not-admit decisions), 91.2% and 87.8%, respectively. The proportion allocated correctly by the LL model was lower for consultants than juniors, whereas the MH model performed equally well for both. In this vignette study, neither model provided any better description of judgments made by consultants or by pediatricians compared to other grades and specialties.

Efficient Non-Linear Idealisations of Aircraft Fuselage Panels in Compression

Aircraft fuselages are complex assemblies of thousands of components and as a result simulation models are highly idealised. In the typical design process, a coarse FE model is used to determine loads within the structure. The size of the model and number of load cases necessitates that only linear static behaviour is considered. This paper reports on the development of a modelling approach to increase the accuracy of the global model, accounting for variations in stiffness due to non-linear structural behaviour. The strategy is based on representing a fuselage sub-section with a single non-linear element. Large portions of fuselage structure are represented by connecting these non-linear elements together to form a framework. The non-linear models are very efficient, reducing computational time significantly

Model selection approaches for non-linear system identification: a review

The identification of non-linear systems using only observed finite datasets has become a mature research area over the last two decades. A class of linear-in-the-parameter models with universal approximation capabilities have been intensively studied and widely used due to the availability of many linear-learning algorithms and their inherent convergence conditions. This article presents a systematic overview of basic research on model selection approaches for linear-in-the-parameter models. One of the fundamental problems in non-linear system identification is to find the minimal model with the best model generalisation performance from observational data only. The important concepts in achieving good model generalisation used in various non-linear system-identification algorithms are first reviewed, including Bayesian parameter regularisation and models selective criteria based on the cross validation and experimental design. A significant advance in machine learning has been the development of the support vector machine as a means for identifying kernel models based on the structural risk minimisation principle. The developments on the convex optimisation-based model construction algorithms including the support vector regression algorithms are outlined. Input selection algorithms and on-line system identification algorithms are also included in this review. Finally, some industrial applications of non-linear models are discussed.

Explaining young mortality

Stochastic modeling of mortality rates focuses on fitting linear models to logarithmically adjusted mortality data from the middle or late ages. Whilst this modeling enables insurers to project mortality rates and hence price mortality products it does not provide good fit for younger aged mortality. Mortality rates below the early 20's are important to model as they give an insight into estimates of the cohort effect for more recent years of birth. It is also important given the cumulative nature of life expectancy to be able to forecast mortality improvements at all ages. When we attempt to fit existing models to a wider age range, 5-89, rather than 20-89 or 50-89, their weaknesses are revealed as the results are not satisfactory. The linear innovations in existing models are not flexible enough to capture the non-linear profile of mortality rates that we see at the lower ages. In this paper we modify an existing 4 factor model of mortality to enable better fitting to a wider age range, and using data from seven developed countries our empirical results show that the proposed model has a better fit to the actual data, is robust, and has good forecasting ability.

Design of neuro-fuzzy models by evolutionary and gradient-based algorithms

All systems found in nature exhibit, with different degrees, a nonlinear behavior. To emulate this behavior, classical systems identification techniques use, typically, linear models, for mathematical simplicity. Models inspired by biological principles (artificial neural networks) and linguistically motivated (fuzzy systems), due to their universal approximation property, are becoming alternatives to classical mathematical models. In systems identification, the design of this type of models is an iterative process, requiring, among other steps, the need to identify the model structure, as well as the estimation of the model parameters. This thesis addresses the applicability of gradient-basis algorithms for the parameter estimation phase, and the use of evolutionary algorithms for model structure selection, for the design of neuro-fuzzy systems, i.e., models that offer the transparency property found in fuzzy systems, but use, for their design, algorithms introduced in the context of neural networks. A new methodology, based on the minimization of the integral of the error, and exploiting the parameter separability property typically found in neuro-fuzzy systems, is proposed for parameter estimation. A recent evolutionary technique (bacterial algorithms), based on the natural phenomenon of microbial evolution, is combined with genetic programming, and the resulting algorithm, bacterial programming, advocated for structure determination. Different versions of this evolutionary technique are combined with gradient-based algorithms, solving problems found in fuzzy and neuro-fuzzy design, namely incorporation of a-priori knowledge, gradient algorithms initialization and model complexity reduction.

Comparing the performance of geostatistical models with additional information from covariates for sewage plume characterization

In this work, kriging with covariates is used to model and map the spatial distribution of salinity measurements gathered by an autonomous underwater vehicle in a sea outfall monitoring campaign aiming to distinguish the effluent plume from the receiving waters and characterize its spatial variability in the vicinity of the discharge. Four different geostatistical linear models for salinity were assumed, where the distance to diffuser, the west-east positioning, and the south-north positioning were used as covariates. Sample variograms were fitted by the Mat`ern models using weighted least squares and maximum likelihood estimation methods as a way to detect eventual discrepancies. Typically, the maximum likelihood method estimated very low ranges which have limited the kriging process. So, at least for these data sets, weighted least squares showed to be the most appropriate estimation method for variogram fitting. The kriged maps show clearly the spatial variation of salinity, and it is possible to identify the effluent plume in the area studied. The results obtained show some guidelines for sewage monitoring if a geostatistical analysis of the data is in mind. It is important to treat properly the existence of anomalous values and to adopt a sampling strategy that includes transects parallel and perpendicular to the effluent dispersion.

Appropriateness criteria for surgery improve clinical outcomes in patients with low back pain and/or sciatica.

STUDY DESIGN: Prospective, controlled, observational outcome study using clinical, radiographic, and patient/physician-based questionnaire data, with patient outcomes at 12 months follow-up. OBJECTIVE: To validate appropriateness criteria for low back surgery. SUMMARY OF BACKGROUND DATA: Most surgical treatment failures are attributed to poor patient selection, but no widely accepted consensus exists on detailed indications for appropriate surgery. METHODS: Appropriateness criteria for low back surgery have been developed by a multispecialty panel using the RAND appropriateness method. Based on panel criteria, a prospective study compared outcomes of patients appropriately and inappropriately treated at a single institution with 12 months follow-up assessment. Included were patients with low back pain and/or sciatica referred to the neurosurgical department. Information about symptoms, neurologic signs, the health-related quality of life (SF-36), disability status (Roland-Morris), and pain intensity (VAS) was assessed at baseline, at 6 months, and at 12 months follow-up. The appropriateness criteria were administered prospectively to each clinical situation and outside of the clinical setting, with the surgeon and patients blinded to the results of the panel decision. The patients were further stratified into 2 groups: appropriate treatment group (ATG) and inappropriate treatment group (ITG). RESULTS: Overall, 398 patients completed all forms at 12 months. Treatment was considered appropriate for 365 participants and inappropriate for 33 participants. The mean improvement in the SF-36 physical component score at 12 months was significantly higher in the ATG (mean: 12.3 points) than in the ITG (mean: 6.8 points) (P = 0.01), as well as the mean improvement in the SF-36 mental component score (ATG mean: 5.0 points; ITG mean: -0.5 points) (P = 0.02). Improvement was also significantly higher in the ATG for the mean VAS back pain (ATG mean: 2.3 points; ITG mean: 0.8 points; P = 0.02) and Roland-Morris disability score (ATG mean: 7.7 points; ITG mean: 4.2 points; P = 0.004). The ATG also had a higher improvement in mean VAS for sciatica (4.0 points) than the ITG (2.8 points), but the difference was not significant (P = 0.08). The SF-36 General Health score declined in both groups after 12 months, however, the decline was worse in the ITG (mean decline: 8.2 points) than in the ATG (mean decline: 1.2 points) (P = 0.04). Overall, in comparison to ITG patients, ATG patients had significantly higher improvement at 12 months, both statistically and clinically. CONCLUSION: In comparison to previously reported literature, our study is the first to assess the utility of appropriateness criteria for low back surgery at 1-year follow-up with multiple outcome dimensions. Our results confirm the hypothesis that application of appropriateness criteria can significantly improve patient outcomes.

Establishing an association between a peri-operative perfusion score system (PerfSCORE) and post-operative patient morbidity/mortality during CPB cardiac surgery.

BACKGROUND: To date, there is no quality assurance program that correlates patient outcome to perfusion service provided during cardiopulmonary bypass (CPB). A score was devised, incorporating objective parameters that would reflect the likelihood to influence patient outcome. The purpose was to create a new method for evaluating the quality of care the perfusionist provides during CPB procedures and to deduce whether it predicts patient morbidity and mortality. METHODS: We analysed 295 consecutive elective patients. We chose 10 parameters: fluid balance, blood transfused, Hct, ACT, PaO2, PaCO2, pH, BE, potassium and CPB time. Distribution analysis was performed using the Shapiro-Wilcoxon test. This made up the PerfSCORE and we tried to find a correlation to mortality rate, patient stay in the ICU and length of mechanical ventilation. Univariate analysis (UA) using linear regression was established for each parameter. Statistical significance was established when p < 0.05. Multivariate analysis (MA) was performed with the same parameters. RESULTS: The mean age was 63.8 +/- 12.6 years with 70% males. There were 180 CABG, 88 valves, and 27 combined CABG/valve procedures. The PerfSCORE of 6.6 +/- 2.4 (0-20), mortality of 2.7% (8/295), CPB time 100 +/- 41 min (19-313), ICU stay 52 +/- 62 hrs (7-564) and mechanical ventilation of 10.5 +/- 14.8 hrs (0-564) was calculated. CPB time, fluid balance, PaO2, PerfSCORE and blood transfused were significantly correlated to mortality (UA, p < 0.05). Also, CPB time, blood transfused and PaO2 were parameters predicting mortality (MA, p < 0.01). Only pH was significantly correlated for predicting ICU stay (UA). Ultrafiltration (UF) and CPB time were significantly correlated (UA, p < 0.01) while UF (p < 0.05) was the only parameter predicting mechanical ventilation duration (MA). CONCLUSIONS: CPB time, blood transfused and PaO2 are independent risk factors of mortality. Fluid balance, blood transfusion, PaO2, PerfSCORE and CPB time are independent parameters for predicting morbidity. PerfSCORE is a quality of perfusion measure that objectively quantifies perfusion performance.

Hierarchical Mixtures of Experts and the EM Algorithm

We present a tree-structured architecture for supervised learning. The statistical model underlying the architecture is a hierarchical mixture model in which both the mixture coefficients and the mixture components are generalized linear models (GLIM's). Learning is treated as a maximum likelihood problem; in particular, we present an Expectation-Maximization (EM) algorithm for adjusting the parameters of the architecture. We also develop an on-line learning algorithm in which the parameters are updated incrementally. Comparative simulation results are presented in the robot dynamics domain.

Time-series regression models to study the short-term effects of environmental factors on health

Time series regression models are especially suitable in epidemiology for evaluating short-term effects of time-varying exposures on health. The problem is that potential for confounding in time series regression is very high. Thus, it is important that trend and seasonality are properly accounted for. Our paper reviews the statistical models commonly used in time-series regression methods, specially allowing for serial correlation, make them potentially useful for selected epidemiological purposes. In particular, we discuss the use of time-series regression for counts using a wide range Generalised Linear Models as well as Generalised Additive Models. In addition, recently critical points in using statistical software for GAM were stressed, and reanalyses of time series data on air pollution and health were performed in order to update already published. Applications are offered through an example on the relationship between asthma emergency admissions and photochemical air pollutants

Mixing compositions and other scales

Theory of compositional data analysis is often focused on the composition only. However in practical applications we often treat a composition together with covariables with some other scale. This contribution systematically gathers and develop statistical tools for this situation. For instance, for the graphical display of the dependence of a composition with a categorical variable, a colored set of ternary diagrams might be a good idea for a first look at the data, but it will fast hide important aspects if the composition has many parts, or it takes extreme values. On the other hand colored scatterplots of ilr components could not be very instructive for the analyst, if the conventional, black-box ilr is used. Thinking on terms of the Euclidean structure of the simplex, we suggest to set up appropriate projections, which on one side show the compositional geometry and on the other side are still comprehensible by a non-expert analyst, readable for all locations and scales of the data. This is e.g. done by defining special balance displays with carefully- selected axes. Following this idea, we need to systematically ask how to display, explore, describe, and test the relation to complementary or explanatory data of categorical, real, ratio or again compositional scales. This contribution shows that it is sufficient to use some basic concepts and very few advanced tools from multivariate statistics (principal covariances, multivariate linear models, trellis or parallel plots, etc.) to build appropriate procedures for all these combinations of scales. This has some fundamental implications in their software implementation, and how might they be taught to analysts not already experts in multivariate analysis