Statistics publications
Permanent URI for this collection
Browse
Browsing Statistics publications by Title
Now showing 1 - 20 of 48
Results Per Page
Sort Options
Item Metadata only A Bayesian approach for optimal reinsurance and investment in a diffusion model(Kluwer Academic Publ, 2012) Zhang, X.; Elliott, R.; Siu, T.A Bayesian adaptive control approach to the combined optimal investment/reinsurance problem of an insurance company is studied. The insurance company invests in a money market and a capital market index with an unknown appreciation rate, or “drift”. Using a Bayesian approach, the unknown drift is described by an unobservable random variable with a known (prior) probability distribution. We assume that the risk process of the company is governed by a diffusion approximation to the compound Poisson risk process. The company also purchases reinsurance. The combined optimal investment/reinsurance problem is formulated as a stochastic optimal control problem with partial observations. We employ filtering theory to transform the problem into one with complete observations. The control problem is then solved by the dynamic programming Hamilton–Jacobi–Bellman (HJB) approach. Semi-analytical solutions are obtained for the exponential utility case.Item Metadata only A conditional likelihood approach to residual maximum likelihood estimation in generalized linear models(Wiley, 1996) Smyth, G.; Verbyla, A.Residual maximum likelihood (REML) estimation is often preferred to maximum likelihood estimation as a method of estimating covariance parameters in linear models because it takes account of the loss of degrees of freedom in estimating the mean and produces unbiased estimating equations for the variance parameters. In this paper it is shown that REML has an exact conditional likelihood interpretation, where the conditioning is on an appropriate sufficient statistic to remove dependence on the nuisance parameters. This interpretation clarifies the motivation for REML and generalizes directly to non-normal models in which there is a low dimensional sufficient statistic for the fitted values. The conditional likelihood is shown to be well defined and to satisfy the properties of a likelihood function, even though this is not generally true when conditioning on statistics which depend on parameters of interest. Using the conditional likelihood representation, the concept of REML is extended to generalized linear models with varying dispersion and canonical link. Explicit calculation of the conditional likelihood is given for the one-way lay-out. A saddlepoint approximation for the conditional likelihood is also derived.Item Metadata only Accounting for natural and extraneous variation in the analysis of field experiments(Springer Verlag, 1997) Gilmour, A.; Cullis, B.; Verbyla, A.We identify three major components of spatial variation in plot errors from field experiments and extend the two-dimensional spatial procedures of Cullis and Gleeson (1991) to account for them. The components are nonstationary, large-scale (global) variation across the field, stationary variation within the trial (natural variation or local trend), and extraneous variation that is often induced by experimental procedures and is predominantly aligned with rows and columns. We present a strategy for identifying a model for the plot errors that uses a trellis plot of residuals, a perspective plot of the sample variogram and, where possible, likelihood ratio tests to identify which components are present. We demonstrate the strategy using two illustrative examples. We conclude that although there is no one model that adequately fits all field experiments, the separable autoregressive model is dominant. However, there is often additional identifiable variation present.Item Metadata only Acute physiology and chronic health evaluation (APACHE II) scoring system recalculated and reconsidered(1996) Moran, J.; Leppard, P.; Finnis, M.; Peisach, R.Item Metadata only AIDS: Modelling and Predicting.(Cambridge University Press., 1996) Solomon, P.Item Metadata only Are causes of death predictable?(Wayne State University Press, 1998) Tallis, George M.; Leppard, Phillip I.; Tallis, G. A.; Hansen, D.A random sample of death records of adult males from 1967 to 1970 was chosen from the South Australian Registry of Births, Deaths, and Marriages. The natural parents of these individuals were identified by cross-reference to birth certificates, and an extensive search was made of the death records for these parents. In this manner random families were selected for which, where possible, the cause of death and length of life of each family member were determined. From the information pertaining to the stated cause of death, each individual was assigned to one of five death categories.Item Open Access Assessment of outcome over a 10-year period of patients admitted to a multidisciplinary adult intensive care unit with haematological and solid tumours(Australian Soc Anaesthetists, 2005) Moran, J.; Solomon, P.; Williams, P.The risk factors for time to mortality, censored at 30 days, of patients admitted to an adult teaching hospital ICU with haematological and solid malignancies were assessed in a retrospective cohort study. Patients, demographics and daily ICU patient data, from admission to day 8, were identified from a prospective computerised database and casenote review in consecutive admissions to ICU with haematological and solid tumours over a 10 year period (1989-99). The cohort, 108 ICU admissions in 89 patients was of mean age (+/-SD) 55+/-14 years; 43% were female. Patient diagnoses were leukaemia (35%), lymphoma (38%) and solid tumours (27%). Median time from hospital to ICU admission was five days (range 0-67). On ICU admission, 50% had septic shock and first day APACHE II score was 28+/-9. Forty six per cent of patients were ventilated. ICU and 30 day mortality were 39% and 54% respectively. Multivariate Cox model predictors (P<0.05), using only ICU admission day data were: Charlson comorbidity index (CCI), time to ICU admission (days) and mechanical ventilation. For daily data (admission through day 8), predictors were: cohort effect (2nd vs 1st five year period); CCI; time to ICU admission (days); APACHE II score and mechanical ventilation. Outcomes were considered appropriate for severity of illness and demonstrated improvement over time. Ventilation was an independent outcome determinant. Controlling for other factors, mortality has improved over time (1st vs 2nd five year period). Analysis restricted to admission data alone may be insensitive to particular covariate effects.Item Metadata only Bayesian Value-at-Risk and expected shortfall forecasting via the asymmetric Laplace distribution(Elsevier Science BV, 2012) Chen, Qian; Gerlach, Richard; Lu, Zudi; School of Mathematical Sciences : StatisticsItem Metadata only Characterizations of continuous distributions and associated goodness of fit tests(Kluwer Academic/plenum Publishers, 2005) Morris, K.; Szynal, D.Item Metadata only Comparative performance of oxygenation indices in ICU patients.(1996) Finnis, M.; Moran, J.; Leppard, P.Item Open Access Does dog or cat ownership lead to increased gastroenteritis in young children in South Australia?(Cambridge Univ Press, 2006) Heyworth, J.; Cutt, H.; Glonek, G.The aim of this study was to investigate the relationship between dog and cat ownership and gastroenteritis in young children. A diary study of 965 children aged 4–6 years living in rural or semi-rural South Australia was undertaken. Data were collected on pet ownership, drinking water and other risk factors for gastroenteritis. Overall 89% of households had pets and dog ownership was more common than cat ownership. The multivariable models for gastroenteritis and pet ownership indicated that living in a household with a dog or cat was associated with a reduced risk of gastroenteritis (adj. OR 0·71, 95% CI 0·55–0·92; OR 0·70, % CI 0·51–0·97 respectively). This paper adds to the evidence that pets are not a major source of gastroenteritis in the home and lends support to the health benefits of pet ownership. However, this must be weighed against the potential negative consequences, such as dog bites, particularly for this age group.Item Metadata only Editorial: Variance components in medical research(1998) Solomon, P.Item Metadata only Effect of mistletoes, Amyema preissi, on host, acacia victoriae, survival(Blackwell Scientific Publications, 1995) Reid, N. C. H.; Stafford Smith, M.; Venables, William N.Item Metadata only Effect of social networks on 10 year survival in very old Australians: the Australian longitudinal study of aging(British Med Journal Publ Group, 2005) Giles, L.; Glonek, G.; Luszcz, M.; Andrews, G.Study objectives: To examine if social networks with children, relatives, friends, and confidants predict survival in older Australians over 10 years after controlling for a range of demographic, health, and lifestyle variables. Design: Prospective longitudinal cohort study (the Australian longitudinal study of aging) Setting: Adelaide, South Australia. Participants: 1477 persons aged 70 years or more living in the community and residential care facilities. Main results: After controlling for a range of demographic, health, and lifestyle variables, greater networks with friends were protective against mortality in the 10 year follow up period. The hazard ratio for participants in the highest tertile of friends networks compared with participants in the lowest group was 0.78 (95%CI 0.65 to 0.92). A smaller effect of greater networks with confidants (hazard ratio = 0.84; 95%CI = 0.71 to 0.98) was seen. The effects of social networks with children and relatives were not significant with respect to survival over the following decade. Conclusions: Survival time may be enhanced by strong social networks. Among older Australians, these may be important in lengthening survival.Item Metadata only Efficient Two-Replicate Resolvable Row-Column Designs(1997) Jarrett, Richard Gentry; Piper, F. C.; Wild, Peter R.Item Metadata only Existence, uniqueness and comparisons for BSDEs in general spaces(Inst Mathematical Statistics, 2012) Cohen, S.; Elliott, R.We present a theory of backward stochastic differential equations in continuous time with an arbitrary filtered probability space. No assumptions are made regarding the left continuity of the filtration, of the predictable quadratic variations of martingales or of the measure integrating the driver. We present conditions for existence and uniqueness of square-integrable solutions, using Lipschitz continuity of the driver. These conditions unite the requirements for existence in continuous and discrete time and allow discrete processes to be embedded with continuous ones.We also present conditions for a comparison theorem and hence construct time consistent nonlinear expectations in these general spaces.Item Metadata only Expression profiling reveals functionally important genes and coordinately regulated signaling pathway genes during in vitro angiogenesis(American Physiological Society, 2005) Hahn, C.; Su, Z.; Drogemuller, C.; Tsykin, A.; Waterman, S.; Brautigan, P.; Yu, S.; Kremmidiotis, G.; Gardner, A.; Solomon, P.; Goodall, G.; Vadas, M.; Gamble, J.Angiogenesis is a complex multicellular process requiring the orchestration of many events including migration, alignment, proliferation, lumen formation, remodeling, and maturation. Such complexity indicates that not only individual genes but also entire signaling pathways will be crucial in angiogenesis. To define an angiogenic blueprint of regulated genes, we utilized our well-characterized three-dimensional collagen gel model of in vitro angiogenesis, in which the majority of cells synchronously progress through defined morphological stages culminating in the formation of capillary tubes. We developed a comprehensive three-tiered approach using microarray analysis, which allowed us to identify genes known to be involved in angiogenesis and genes hitherto unlinked to angiogenesis as well as novel genes and has proven especially useful for genes where the magnitude of change is small. Of interest is the ability to recognize complete signaling pathways that are regulated and genes clustering into ontological groups implicating the functional importance of particular processes. We have shown that consecutive members of the mitogen-activated protein kinase and leukemia inhibitory factor signaling pathways are altered at the mRNA level during in vitro angiogenesis. Thus, at least for the mitogen-activated protein kinase pathway, mRNA changes as well as the phosphorylation changes of these gene products may be important in the control of blood vessel morphogenesis. Furthermore, in this study, we demonstrated the power of virtual Northern blot analysis, as an alternative to quantitative RT-PCR, for measuring the magnitudes of differential gene expression.Item Metadata only Extending mixtures of factor models using the restricted multivariate skew-normal distribution(Elsevier, 2016) Lin, T.-I.; McLachlan, G.J.; Lee, S.X.The mixture of factor analyzers (MFA) model provides a powerful tool for analyzing high-dimensional data as it can reduce the number of free parameters through its factor-analytic representation of the component covariance matrices. This paper extends the MFA model to incorporate a restricted version of the multivariate skew-normal distribution for the latent component factors, called mixtures of skew-normal factor analyzers (MSNFA). The proposed MSNFA model allows us to relax the need of the normality assumption for the latent factors in order to accommodate skewness in the observed data. The MSNFA model thus provides an approach to model-based density estimation and clustering of high-dimensional data exhibiting asymmetric characteristics. A computationally feasible Expectation Conditional Maximization (ECM) algorithm is developed for computing the maximum likelihood estimates of model parameters. The potential of the proposed methodology is exemplified using both real and simulated data.Item Metadata only Factors affecting survival after endovascular aneurysm repair: results from a population based audit(W B Saunders Co Ltd, 2007) Boult, M.; Maddern, G.; Barnes, M.; Fitridge, R.Objectives: To determine the effect of pre-operative factors on mid-term survival of patients enrolled in an Australian audit of endovascular aneurysm repair (EVAR). Design: Prospective longitudinal national register (audit) of patients undergoing EVAR. Methods: 961 individuals who had elective or semi-urgent EVAR of abdominal aortic aneurysms were enrolled in the audit between November 1999 and May 2001. Data was contributed by 81 surgeons from 64 hospitals. Kaplan-Meier survival analysis was used to determine survival rates and factors significantly influencing survival. Parametric survival analysis with log-exponential distribution was used to estimate expected 3 and 5 year survival for different ages, ASA, creatinine and aneurysm sizes. Results: Overall survival was 93% at 1 year, 80% at 3 years and 67% at five years. Survival rates were found to be statistically associated with ASA, age, aneurysm size and creatinine levels. ASA has the largest effect. Five year survival rates for aneurysms ≥65 mm and <55 mm were 54% and 76% respectively. Pre-operative creatinine levels ≥ 160 μmol/L lowered the survival rate from 71% to 40%. Conclusions: Survival for EVAR patients is strongly correlated with a number of pre-operative factors. This survival analysis provides a useful decision-making tool for surgeons particularly for individuals with smaller aneurysms.Item Metadata only Genotype by Environment Variance Heterogeneity in a Two-Stage Analysis(1997) Frensham, A.; Cullis, B.; Verbyla, A.
- «
- 1 (current)
- 2
- 3
- »