Modelling inpatient length of stay by a hierarchical mixture regression via the EM algorithm

The modelling of inpatient length of stay (LOS) has important implications in health care studies. Finite mixture distributions are usually used to model the heterogeneous LOS distribution, due to a certain proportion of patients sustaining-a longer stay. However, the morbidity data are collected from hospitals, observations clustered within the same hospital are often correlated. The generalized linear mixed model approach is adopted to accommodate the inherent correlation via unobservable random effects. An EM algorithm is developed to obtain residual maximum quasi-likelihood estimation. The proposed hierarchical mixture regression approach enables the identification and assessment of factors influencing the long-stay proportion and the LOS for the long-stay patient subgroup. A neonatal LOS data set is used for illustration, (C) 2003 Elsevier Science Ltd. All rights reserved.

Hierarchical models for 2D presence/absence data having ambiguous zeroes: With a biogeographical case study on dingo behaviour

This dissertation is primarily an applied statistical modelling investigation, motivated by a case study comprising real data and real questions. Theoretical questions on modelling and computation of normalization constants arose from pursuit of these data analytic questions. The essence of the thesis can be described as follows. Consider binary data observed on a two-dimensional lattice. A common problem with such data is the ambiguity of zeroes recorded. These may represent zero response given some threshold (presence) or that the threshold has not been triggered (absence). Suppose that the researcher wishes to estimate the effects of covariates on the binary responses, whilst taking into account underlying spatial variation, which is itself of some interest. This situation arises in many contexts and the dingo, cypress and toad case studies described in the motivation chapter are examples of this. Two main approaches to modelling and inference are investigated in this thesis. The first is frequentist and based on generalized linear models, with spatial variation modelled by using a block structure or by smoothing the residuals spatially. The EM algorithm can be used to obtain point estimates, coupled with bootstrapping or asymptotic MLE estimates for standard errors. The second approach is Bayesian and based on a three- or four-tier hierarchical model, comprising a logistic regression with covariates for the data layer, a binary Markov Random field (MRF) for the underlying spatial process, and suitable priors for parameters in these main models. The three-parameter autologistic model is a particular MRF of interest. Markov chain Monte Carlo (MCMC) methods comprising hybrid Metropolis/Gibbs samplers is suitable for computation in this situation. Model performance can be gauged by MCMC diagnostics. Model choice can be assessed by incorporating another tier in the modelling hierarchy. This requires evaluation of a normalization constant, a notoriously difficult problem. Difficulty with estimating the normalization constant for the MRF can be overcome by using a path integral approach, although this is a highly computationally intensive method. Different methods of estimating ratios of normalization constants (N Cs) are investigated, including importance sampling Monte Carlo (ISMC), dependent Monte Carlo based on MCMC simulations (MCMC), and reverse logistic regression (RLR). I develop an idea present though not fully developed in the literature, and propose the Integrated mean canonical statistic (IMCS) method for estimating log NC ratios for binary MRFs. The IMCS method falls within the framework of the newly identified path sampling methods of Gelman & Meng (1998) and outperforms ISMC, MCMC and RLR. It also does not rely on simplifying assumptions, such as ignoring spatio-temporal dependence in the process. A thorough investigation is made of the application of IMCS to the three-parameter Autologistic model. This work introduces background computations required for the full implementation of the four-tier model in Chapter 7. Two different extensions of the three-tier model to a four-tier version are investigated. The first extension incorporates temporal dependence in the underlying spatio-temporal process. The second extensions allows the successes and failures in the data layer to depend on time. The MCMC computational method is extended to incorporate the extra layer. A major contribution of the thesis is the development of a fully Bayesian approach to inference for these hierarchical models for the first time. Note: The author of this thesis has agreed to make it open access but invites people downloading the thesis to send her an email via the 'Contact Author' function.

Modelling the water, energy and economic nexus

The mining industry faces three long term strategic risks in relation to its water and energy use: 1) securing enough water and energy to meet increased production; 2) reducing water use, energy consumption and emissions due to social, environmental and economic pressures; and 3) understanding the links between water and energy, so that an improvement in one area does not create an adverse effect in another. This project helps the industry analyse these risks by creating a hierarchical systems model (HSM) that represents the water and energy interactions on a sub-site, site and regional scales; which is coupled with a flexible risk framework. The HSM consists of: components that represent sources of water and energy; activities that use water and energy and off-site destinations of water and produced emissions. It can also represent more complex components on a site, with inbuilt examples including tailings dams and water treatment plants. The HSM also allows multiple sites and other infrastructure to be connected together to explore regional water and energy interactions. By representing water and energy as a single interconnected system the HSM can explore tradeoffs and synergies. For example, on a synthetic case study, which represents a typical site, simulations suggested that while a synergy in terms of water use and energy use could be made when chemical additives were used to enhance dust suppression, there were trade-offs when either thickened tailings or dry processing were used. On a regional scale, the HSM was used to simulate various scenarios, including: mines only withdrawing water when needed; achieving economics-of-scale through use of a single centralised treatment plant rather than smaller decentralised treatment plants; and capturing of fugitive emissions for energy generation. The HSM also includes an integrated risk framework for interpreting model output, so that onsite and off-site impacts of various water and energy management strategies can be compared in a managerial context. The case studies in this report explored company, social and environmental risks for scenarios of regional water scarcity, unregulated saline discharge, and the use of plantation forestry to offset carbon emissions. The HSM was able to represent the non-linear causal relationship at the regional scale, such as the forestry scheme offsetting a small percentage of carbon emissions but causing severe regional water shortages. The HSM software developed in this project will be released as an open source tool to allow industry personnel to easily and inexpensively quantify and explore the links between water use, energy use, and carbon emissions. The tool can be easily adapted to represent specific sites or regions. Case studies conducted in this project highlighted the potential complexity of these links between water, energy, and carbon emissions, as well as the significance of the cumulative effects of these links over time. A deeper understanding of these links is vital for the mining industry in order to progress to more sustainable operations, and the HSM provides an accessible, robust framework for investigating these links.

Predicting health programme participation: A gravity‐based, hierarchical modelling approach

Statistical analyses of health program participation seek to address a number of objectives compatible with the evaluation of demand for current resources. In this spirit, a spatial hierarchical model is developed for disentangling patterns in participation at the small area level, as a function of population-based demand and additional variation. For the former, a constrained gravity model is proposed to quantify factors associated with spatial choice and account for competition effects, for programs delivered by multiple clinics. The implications of gravity model misspecification within a mixed effects framework are also explored. The proposed model is applied to participation data from a no-fee mammography program in Brisbane, Australia. Attention is paid to the interpretation of various model outputs and their relevance for public health policy.

Hierarchical modelling of acclimatory processesstar, open

The term acclimation has been used with several connotations in the field of acclimatory physiology. An attempt has been made, in this paper, to define precisely the term “acclimation” for effective modelling of acclimatory processes. Acclimation is defined with respect to a specific variable, as cumulative experience gained by the organism when subjected to a step change in the environment. Experimental observations on a large number of variables in animals exposed to sustained stress, show that after initial deviation from the basal value (defined as “growth”), the variables tend to return to basal levels (defined as “decay”). This forms the basis for modelling biological responses in terms of their growth and decay. Hierarchical systems theory as presented by Mesarovic, Macko & Takahara (1970) facilitates modelling of complex and partially characterized systems. This theory, in conjunction with “growth-decay” analysis of biological variables, is used to model temperature regulating system in animals exposed to cold. This approach appears to be applicable at all levels of biological organization. Regulation of hormonal activity which forms a part of the temperature regulating system, and the relationship of the latter with the “energy” system of the animal of which it forms a part, are also effectively modelled by this approach. It is believed that this systematic approach would eliminate much of the current circular thinking in the area of acclimatory physiology.

Scalable non-linear Support Vector Machine using hierarchical clustering

This paper discusses a method for scaling SVM with Gaussian kernel function to handle large data sets by using a selective sampling strategy for the training set. It employs a scalable hierarchical clustering algorithm to construct cluster indexing structures of the training data in the kernel induced feature space. These are then used for selective sampling of the training data for SVM to impart scalability to the training process. Empirical studies made on real world data sets show that the proposed strategy performs well on large data sets.