No description available.

Functional data is ubiquitous in modern science, technology and medicine. An example, which motivates the work in this thesis, is an experiment in tribology to investigate wear in automotive transmission. The research in this thesis provides methods for the design of experiments when the response is assumed to be a realisation of a smooth function. In the course of the research, two areas were investigated: designs for local linear smoothers and designs for discriminating between two functional linear models. Designs that are optimal for minimising the prediction variance of a smooth function were found across an interval using two kernel smoothing methods: local linear regression and Gasser and Muller estimation. The values of the locality parameter and run size were shown to affect the optimal design. Optimal designs for best prediction using local linear regression were applied to the tribology experiment. A compound optimality criterion is proposed which is a weighted average of the integrated prediction variance and the inverse of the trace of the smoothing matrix using the Gasser and Muller estimator. The complexity of the model to be fitted was shown to influence the selection of optimal design points. The robustness of these optimal designs to misspecification of the kernel function for the compound criterion was also critically assessed. A criterion and method for finding T-optimal designs was developed for discriminating between two competing functional linear models. It was proved that the choice of optimal design is independent of the parameter values when discriminating between two nested functional linear models that differ by only one term. The performance of T-optimal designs was evaluated in simulation studies which calculated the power of the test for assessing the fit of one model using data generated from the competing model.

A central objective of current systems biology research is explaining the interactions amongst components in biopathways. A standard approach is to view a biopathway as a network of biochemical reactions, which is modelled as a system of ordinary differential equations (ODEs). Conventional inference methods typically rely on searching the space of parameter values, and at each candidate, numerically solving the ODEs and comparing the output with that observed. After choosing an appropriate noise model, the form of the likelihood is defined, and a measure of similarity between the data signals and the signals described by the current set of ODE parameters can be calculated. This process is repeated, as part of either an iterative optimisation scheme or sampling procedure in order to estimate the parameters. However, the computational costs involved with repeatedly numerically solving the ODEs are usually high. Several authors have adopted approaches based on gradient matching, aiming to reduce this computational complexity. These approaches are based on the following two-step procedure. At the first step, interpolation is used to smooth the time series data, in order to avoid modelling noisy observations; in a second step, the kinetic parameters of the ODEs are either optimised or sampled, whilst minimising some metric measuring the difference between the slopes of the tangents to the interpolants, and the parameter-dependent time derivative from the ODEs. In this fashion, the ODEs never have to be numerically integrated, and the problem of inferring the typically unknown initial conditions of the system is removed, as it is not required for matching gradients. A downside to this two-step scheme is that the results of parameter inference are critically dependent on the quality of the initial interpolant. Alternatively, the ODEs can be allowed to regularise the interpolant and it has been demonstrated that it significantly improves the parameter inference accuracy and robustness with respect to noise. This thesis extends and develops methods of gradient matching for parameter inference and model selection in ODE systems in a systems biology context.

This Ph.D. thesis contains 3 essays in international finance with a focus on foreign exchange market from the perspectives of empirical asset pricing (Chapter 2 and Chapter 3), forecasting and market microstructure (Chapter 4). In Chapter 2, I derive the position-unwinding likelihood indicator for currency carry trade portfolios in the option pricing model, and show that it represents the systematic crash risk associated with global liquidity imbalances and also is able to price the cross-section of global currency, sovereign bond, and equity portfolios; I also explore the currency option-implied sovereign default risk in Merton’s framework, and link the sovereign CDS-implied credit risk premia to currency excess returns that it prices the cross section of currency carry, momentum, and volatility risk premium portfolios. In Chapter 3, I investigate the factor structure in currency market and identify three important properties of global currencies – overvalued (undervalued) currencies with respect to equilibrium exchange rates tend to be crash sensitive (insensitive) measured by copula lower tail dependence, relatively cheap (expensive) to hedge in terms of volatility risk premium, and exposed to high (low) speculative propensity gauged by skew risk premium. I further reveal that these three characteristics have rich asset pricing and asset allocation implications, e.g. striking crash-neutral and diversification benefits for portfolio optimization and risk management purposes. In Chapter 4, I examine the term structure of exchange rate predictability by return decomposition, incorporate common latent factors across a range of investment horizons into the exchange rate dynamics with a broad set of predictors, and handle both parameter uncertainty and model uncertainty. I demonstrate the time-varying term-structural effect and model disagreement effect of exchange rate determinants and the projections of predictive information over the term structure, and utilize the time-variation in the probability weighting from dynamic model averaging to identify the scapegoat drivers of customer order flows. I further comprehensively evaluate both statistical and economic significance of the model allowing for a full spectrum of currency investment management, and find that the model generates substantial performance fees.

This thesis develops statistical methodology for disease mapping, an increasingly important field of spatial epidemiology. Disease mapping has applications in public health by allowing for identification of areas which are at high risk of particular health problems. Such approaches are generally based on areal data, which involves partitioning the study region into a set of non-overlapping areal units and recording counts of disease cases within each areal unit. The majority of approaches assume a spatially smooth risk surface, but this may not be realistic, and there has been recent interest in developing methodology which allows for discontinuities in this structure. This can be done by identifying clusters of areal units with similar disease risks, and allowing for discontinuities between these clusters. The work presented in this thesis develops models to identify such clusters and also estimate disease risk. Three Bayesian hierarchical models are proposed; the first two are based on spatial data at a single time point, while the third extends into the spatio-temporal domain by modelling across multiple time points. Each model is applied to respiratory hospital admission data from the Greater Glasgow and Clyde Health Board area in order to identify clusters which have high disease risk.

Notwithstanding the proven achievements of the New-Keynesian research programme, the models currently used for monetary policy analysis rely on two assumptions that are often taken for granted. One is the balanced growth path property, which has generally been an accurate description of the US and other advanced economies. The other assumption concerns the small volatility of shocks that enables the researcher to approximate the solution of the original model locally. In the past decade, however, emerging economies such as China, Brazil, the Czech Republic or Poland have experienced persistent growth rates of GDP per capita that have been well above the corresponding levels in the euro area or the US. But how should monetary policy respond to an ongoing real convergence process which precisely differentiates emerging from advanced economies? The first part of the thesis aims to answer this question in the context of economies also bound to become future members of the euro area. Owing to the long-term institutional commitment to satisfy the Maastricht convergence criteria during the ERM-II mechanism, policy makers in Central Europe face the additional responsibility of managing the tension between nominal and real convergence. For instance, the Balassa-Samuelson hypothesis postulates an empirically relevant reason as to why countries engaged in a catching-up process might experience a higher inflation rate brought about by the increase in the relative price of services. Motivated by the stylised facts of macroeconomic dynamics in the Czech Republic, a country we take as representative for the whole region, Chapter 1 develops a stylised SOE model with nominal rigidities that is subject to asymmetric productivity growth shocks affecting the traded and nontraded sectors. Relative to the existing literature analysing optimal monetary policy under commitment in Balassa-Samuelson type of macroeconomic environments, the model we propose differentiates itself in that it allows for endogenous current account fluctuations and uncorrected steady state distortions. These modifications result in richer dynamics, which are shaped by the possibility to influence the terms of trade in one’s favour and the presence of monopoly power in product markets. In setting up the welfare maximising interest rate responses, the optimal plan trades off conflicting inflationary and deflationary incentives stemming from the existence of the above externalities. Whereas the first chapter focuses on the methods and assumptions needed to detrend the nonstationary model, the second chapter examines the optimal monetary policy stance under real convergence in two different market structures. The simulations reveal that the specific policy recommendations depend on the degree of substitutability between domestic and foreign goods, a parameter which also alters the strength of the wealth effects driving consumption responses. When monopolistic competition in the traded sector is assumed, the Ramsey interest rate plan is countercyclical. Owing to a cancellation of the terms of trade externality, the predictions are however reversed under perfect competition. This is because the incentive to stimulate production away from the inefficient steady state level becomes dominant. Additionally, the study conducts an extensive welfare analysis through which the effectiveness of inflation targeting and exchange rate peg regimes is assessed relative to the Ramsey plan. It is shown that policies achieving appropriate measures of price stability robustly deliver higher conditional welfare during a catching-up process. The analysis is suggestively complemented with policy experiments that are relevant to the ERM-II period, such as the Maastricht constrained optimal plan, its welfare costs and the welfare-maximising choice of a central parity at which the nominal exchange rate should be fixed. The final part of the thesis examines the macroeconomic costs of euro adoption in Emerging Europe, conditional on the EMU membership eventuality. Inspired from the Optimum Currency Areas literature, the research conducted in the third chapter investigates the circumstances when the decisions made by the ECB would correspond to the domestic optimal interest rate responses. The empirical work looks at the structural alignment and the degree of business cycle synchronisation between prospective and current members of the single currency area, modelled suggestively as the Czech and Austrian economies. A rich SOE model with incomplete markets and trade in intermediate inputs is developed in this sense, whose core structure is similar to Kollmann (2001). Relative to the original framework, we augment its shock structure and enrich the dynamics by incorporating external habit formation and partial indexation in the Calvo adjustment rules for prices and wages. The state-space representation of the DSGE model is taken to data and the set of random parameters is estimated using Bayesian techniques. The comparative analysis reveals that most structural parameters are not very far from each other, suggesting that a moderate degree of structural convergence has been achieved by the emerging economy. The costs of losing monetary policy sovereignty are further assessed by employing a battery of tests, which include impulse response analyses and historical decompositions of output and inflation. While confirming previous SVAR evidence, the results suggest that the propagation mechanisms of monetary policy, productivity and demand shocks are remarkably similar across the two economies. In contrast, the analysis also indicates considerable asymmetries of the sources of fluctuations, which were more volatile and largely idiosyncratic in the Czech Republic. The low degree of business cycle synchronisation suggests that coping with euro area interest rates on a permanent basis is likely to be painful.

Several human genetic diseases, including myotonic dystrophy type 1 (DM1) and Huntington disease (HD), are associated with inheriting an abnormally large unstable DNA simple sequence tandem repeat. These sequences mutate, by changing the number of repeats, many times during the lifetime of those affected, with a bias towards expansion. High repeat numbers are associated with early onset and disease severity. The presence of somatic instability compromises attempts to measure intergenerational repeat dynamics and infer genotype-phenotype relationships. Modelling the progression of repeat length throughout the lifetime of individuals has potential for improving prognostic information as well as providing a deeper understanding of the underlying biological process. Dr Fernando Morales, Dr Anneli Cooper and others from the Monckton lab have characterised more than 25,000 de novo somatic mutations from a large cohort of DM1 patients using single-molecule polymerase chain reaction (SM-PCR). This rich dataset enables us to fully quantify levels of somatic instability across a representative DM1 population for the first time. We establish the relationship between inherited or progenitor allele length, age at sampling and levels of somatic instability using linear regression analysis. We show that the estimated progenitor allele length genotype is significantly better than modal repeat length (the current clinical standard) at predicting age of onset and this novel genotype is the major modifier of the age of onset phenotype. Further we show that somatic variation (adjusted for estimated progenitor allele length and age at sampling) is also a modifier of the age of onset phenotype. Several families form the large cohort, and we find that the level of somatic instability is highly heritable, implying a role for individual-specific trans-acting genetic modifiers. We develop new mathematical models, the main focus of this thesis, by modifying a previously proposed stochastic birth process to incorporate possible contraction. A Bayesian likelihood approach is used as the basis for inference and parameter estimation. We use model comparison analysis to reveal, for the first time, that the expansion bias observed in the distributions of repeat lengths is likely to be the cumulative effect of many expansion and contraction events. We predict that mutation events can occur as frequently as every other day, which matches the timing of regular cell activities such as DNA repair and transcription, but not DNA replication. Mutation rates estimated under the models described above are lower than expected among individuals with inherited repeat lengths less than 100 CTGs, suggesting that these rates may be suppressed at the lower end of the disease causing range. We propose that a length-specific effect may be operating within this range and test this hypothesis by introducing such an effect into the model. To calibrate this extended model, we use blood DNA data from DM1 individuals with small alleles (inherited repeat lengths less than 100 CTGs) and buccal DNA from HD individuals who almost always have inherited repeat lengths less than 100 CAGs. These datasets comprise single DNA molecules sized using SM-PCR. We find statistical support for a general length-specific effect which suppresses mutational rates among the smaller alleles and gives rise to a distinctive pattern in the repeat length distributions. In a novel application of this new model, fitted to a large cohort of DM1 individuals, we also show that this distinctive pattern may help identify individuals whose effective repeat length, with regards to somatic instability, is less than their actual repeat length. A plausible explanation for this distinction is that the expanded repeat tract is compromised by interruptions or other unusual features. For these individuals, we estimate the effective repeat length of their expanded repeat tracts and contribute to the on-going discussion about the effect of interruptions on phenotype. The interpretation of the levels of somatic instability in many of the affected tissues in the triplet repeat diseases is hindered by complex cell compositions. We extend our model to two cell populations whose repeat lengths have different rates of mutation (fast and slow). Swami et al. have recently characterised repeat length distributions in end stage HD brain. Applying our model, we infer for each frontal cortex HD dataset the likely relative weight of these cell populations and their corresponding contribution towards somatic variation. By comparison with data from laser captured single cells we conclude that the neuronal repeat lengths most likely mutate at a higher rate than glial repeat lengths, explaining the characteristic skewed distributions observed in mixed cell tissue from the brain. We confirm that individual-specific mutation rates in neurons are, in addition to the inherited repeat length, a modifier of age of onset. Our results support a model of disease progression where individuals with the same inherited repeat length may reach age of onset, as much as 30 years earlier, because of greater somatic expansions underpinned by higher mutational rates. Therapies aimed at reducing somatic expansions would therefore have considerable benefits with regard to extending the age of onset. Currently clinical diagnosis of DM1 is based on a measure of repeat length from blood cells, but variance in modal length only accounts for between 20 - 40% of the variance in age of onset and, therefore, is not a an accurate predictive tool. We show that in principle progenitor allele length improves the inverse correlation with age of onset over the traditional model length measure. We make use of second blood samples that are now available from 40 DM1 individuals. We show that inherited repeat length and the mutation rates underlying repeat length instability in blood, inferred from samples at two time points rather than one, are better predictors of age of onset than the traditional modal length measure. Our results are a step towards providing better prognostic information for DM1 individuals and their families. They should also lead to better predictions for drug/therapy response, which is emerging as key to successful clinical trials. Microsatellites are another type of tandem repeat found in the genome with high levels of intergenerational and somatic mutation. Differences between individuals make microsatellites very useful biomarkers and they have many applications in forensics and medicine. As well as a general application to other expanded repeat diseases, the mathematical models developed here could be used to better understand instability at other mutational hotspots such as microsatellites.

This thesis explores the role of trade and trade liberalization policies on Tanzanian economy with special focus on the performance of agricultural sector. In terms of methodology, we first use parametric and non-parametric tests to evaluate the impact of liberalization policies on the growth rate of exports. Secondly, we use ordinary least square and instrumental variable to test the “inverse relationship hypothesis” and then we estimate the effect of liberalization on land productivity. We also extend this analysis to Uganda in order to ascertain whether similar findings could be replicated in other developing countries. Thirdly, we employ the co-integration technique to evaluate the effects of openness on economic growth. The parametric and non-parametric tests shows that: despite the marked variation in the composition of traditional exports especially during the late 1990s; largely from coffee and cotton to cashewnuts and tobacco, the contribution of trade liberalization in fostering export growth is rather weak. Second, although the volume of food crops during the post reform period is much higher than before the reforms, there are no symptoms of increased growth overtime. The empirical evidence from econometric analysis shows the existence of diminishing returns to land in the agricultural sector. On the other hand, the impact of trade liberalization on land productivity is mixed; while in some traditional exports its impact is negative and significant, in others the impact is positive but not significant. Contrary to the conventional wisdom as documented in the traditional theories of comparative advantage, the problem with Tanzanian agriculture is not related to the land size but low productivity. Interestingly, these results are also replicated in the Ugandan case. The cointegration analysis shows that the share of trade to GDP is negatively correlated with economic growth. In general, the contribution of this thesis has wider implications in the development policy, at least for the case of Tanzania and other developing countries. First, trade liberalization policies are counterproductive unless diminishing returns to land is squarely addressed. Secondly, the existence of diminishing returns to land is incompatible with the simple prediction of the theory of comparative advantage. The presumption behind trade liberalization is that specialization according to the “comparative advantage” doctrine would inevitably enhance increased productivity (i.e., efficiency). Our results do not conform to this presumption. Third, diminishing returns means that as production increases with international specialization, every additional unit of commodity produced would be more expensive to produce. Fourth, the persistence of diminishing returns to land is incompatible with poverty reduction.

Background Stroke is the leading cause of long-term neurological disability in adults and the third most common cause of death in Britain. It is well known that in addition to the patient characteristics of age and severity, the treatment a stroke patient receives in hospital signifcantly affects outcome. The effectiveness of complex service interventions, how the benefits of these interventions are achieved and the economic impact of different types of service delivery were explored. Methods The Stroke Unit Trialists' Collaboration systematic review was updated and currently contains 31 clinical trials (6936 subjects). The aims were explored using various basic frequentist and Bayesian meta-analysis techniques as well as more complex meta-analysis ideas. These more complex ideas include: meta-regression where covariate information is incorporated into the model; and network meta-analysis where direct and indirect information is used in a mixed treatment comparisons model while also incorporating covariate information. Results Organised inpatient (stroke unit) care showed reductions in death, death or dependency and death or institutional care compared to general medical wards. Stroke unit care appears to reduce the risk of adverse outcomes through prevention and treatment of complications. Acute, comprehensive and rehabilitation stroke unit care appeared to be most effective and acute stroke unit care appeared to be the most cost-effective. However, acute followed by rehabilitation stroke unit care, if required, appears to be the most cost-effective pathway of care compared to the other pathways analysed. Discussion Future research should focus on rehabilitation, acute and comprehensive systems of inpatient care, and explore the best ways of preventing and managing specific complications. Effort should be made to make individual patient data and information on the care pathway of a stroke patient available for meta-analysis.

This is a study of student attrition amongst full-time, first year undergraduates at the University of Glasgow during the 1999-2000 academic session. The thesis contains an initial assessment of the importance of research in this area (Chapter 1), followed by a review of the literature, focusing in particular on the theories and explanations of student attrition that have been advanced by other authors (Chapter 2), and on appropriate research methodologies and data collection techniques (Chapter 3). The investigation then progresses through a succession of different empirical and data-analytic phases. Because of his function within the organisation, the author had uniquely good access to the student records system maintained centrally by the University. This made it practical to sift through this information in such a way as to determine first the simple concomitances of retention (Chapter 4), and then to use it in a more sophisticated manner to develop logistic regression models of retention (Chapters 5 and 8). The challenge was then to decide which new, additional data should be gathered in order to improve upon these quantitative models. The solutions were found partly by recourse to some focus group work with students and staff (Chapter 6). This resulted in two questionnaires being developed to discover students’ attitudes believed to be relevant to retention (Chapter 6). The first survey instrument was administered to all first-year students as part of the matriculation process. The other was completed on-line in the course of the session as an adjunct to the IT Induction Programme for all first-year students. Chapter 10 contains the first outcomes of the attempt to improve the logistic regression models described in Chapter 5 by the introduction of attitudinal constructs, first on their own, and then in combination with the original background and prior academic characteristics in order to model summer retention. The amount of data available in this study is considerable and, consequently, some large-sample structural equation techniques were then used to develop some new, more comprehensive models of retention (Chapter 11). These are more informative, demonstrating how trade-offs can occur between different variables in an overall model of retention, and identifying particular areas where practical policy interventions are likely to be successful in ameliorating student attrition. It is demonstrated that summer retention is affected in roughly equal measure by academic and non-academic factors. On the academic side, it is shown that extra effort and additional academic help and feedback can benefit those students having relatively low entry point scores, for example. Social integration, at least in moderation, is beneficial, and it is positively influenced by living in university accommodation. However, various extraneous problems harm retention through the mediating variables of social integration and commitment. The models have a temporal dimension, and it is argued that students’ attitudes whilst on course owe their origins to those detected at the time of matriculation and, ultimately, back to levels of family support.