We discuss efficient Bayesian estimation of dynamic covariance matrices in multivariate time series through a factor stochastic volatility model. In particular, we propose two interweaving strategies (Yu and Meng, Journal of Computational and Graphical Statistics, 20(3), 531-570, 2011) to substantially accelerate convergence and mixing of standard MCMC approaches. Similar to marginal data augmentation techniques, the proposed acceleration procedures exploit non-identifiability issues which frequently arise in factor models. Our new interweaving strategies are easy to implement and come at almost no extra computational cost; nevertheless, they can boost estimation efficiency by several orders of magnitude as is shown in extensive simulation studies. To conclude, the application of our algorithm to a 26-dimensional exchange rate data set illustrates the superior performance of the new approach for real-world data. / Series: Research Report Series / Department of Statistics and Mathematics

No description available.

Clinical drug trials are costly and time-consuming. Bayesian methods alleviate the inefficiencies in the testing process while providing user-friendly probabilistic inference and predictions from the sampled posterior distributions, saving resources, time, and money. We propose a dynamic linear model to estimate the mean response at each dose level, borrowing strength across dose levels. Our model permits nonmonotonicity of the dose-response relationship, facilitating precise modeling of a wider array of dose-response relationships (including the possibility of toxicity). In addition, we incorporate an adaptive approach to the design of the clinical trial, which allows for interim decisions and assignment to doses based on dose-response uncertainty and dose efficacy. The interim decisions we consider are stopping early for success and stopping early for futility, allowing for patient and time savings in the drug development process. These methods complement current clinical trial design research.

This paper directs interest on country-specific labour market discrimination Roma may suffer in South East Europe. The study lies in the tradition of statistical Blinder-Oaxaca decomposition analysis. We use microdata from UNDP's 2004 survey of Roma minorities, and apply a Bayesian approach, proposed by Keith and LeSage (2004), for the decomposition analysis of wage differentials. This approach is based on a robust Bayesian heteroscedastic linear regression model in conjunction with Markov Chain Monte Carlo (MCMC) estimation. The results obtained indicate the presence of labour market discrimination in Albania and Kosovo, but point to its absence in Bulgaria, Croatia, and Serbia. (authors' abstract)

Dealing with missing data in spatio-temporal time series constitutes important branch of general missing data problem. Since the statistical properties of time-dependent data characterized by sequentiality of observations then any interruption of consecutiveness in time series will cause severe problems. In order to make reliable analyses in this case missing data must be handled cautiously without disturbing the series statistical properties, mainly as temporal and spatial dependencies. In this study we aimed to compare several imputation methods for the appropriate completion of missing values of the spatio-temporal meteorological time series. For this purpose, several missing imputation methods are assessed on their imputation performances for artificially created missing data in monthly total precipitation and monthly mean temperature series which are obtained from the climate stations of Turkish State Meteorological Service. Artificially created missing data are estimated by using six methods. Single Arithmetic Average (SAA), Normal Ratio (NR) and NR Weighted with Correlations (NRWC) are the three simple methods used in the study. On the other hand, we used two computational intensive methods for missing data imputation which are called Multi Layer Perceptron type Neural Network (MLPNN) and Monte Carlo Markov Chain based on Expectation-Maximization Algorithm (EM-MCMC). In addition to these, we propose a modification in the EM-MCMC method in which results of simple imputation methods are used as auxiliary variables. Beside the using accuracy measure based on squared errors we proposed Correlation Dimension (CD) technique for appropriate evaluation of imputation performances which is also important subject of Nonlinear Dynamic Time Series Analysis.

This thesis consists of three chapters on economic and econometric applications of Bayesian parameter estimation and model comparison. The first two chapters study the incidental parameter problem mainly under a linear autoregressive (AR) panel data model with fixed effect. The first chapter investigates the problem from a model comparison perspective. The major finding in the first chapter is that consistency in parameter estimation and model selection are interrelated. The reparameterization of the fixed effect parameter proposed by Lancaster (2002) may not provide a valid solution to the incidental parameter problem if the wrong set of exogenous regressors are included. To estimate the model consistently and to measure its goodness of fit, the Bayes factor is found to be more preferable for model comparson than the Bayesian information criterion based on the biased maximum likelihood estimates. When the model uncertainty is substantial, Bayesian model averaging is recommended. The method is applied to study the relationship between financial development and economic growth. The second chapter proposes a correction function approach to solve the incidental parameter problem. It is discovered that the correction function exists for the linear AR panel model of order p when the model is stationary with strictly exogenous regressors. MCMC algorithms are developed for parameter estimation and to calculate the Bayes factor for model comparison. The last chapter studies how stock return's predictability and model uncertainty affect a rational buy-and-hold investor's decision to allocate her wealth for different lengths of investment horizons in the UK market. The FTSE All-Share Index is treated as the risky asset, and the UK Treasury bill as the riskless asset in forming the investor's portfolio. Bayesian methods are employed to identify the most powerful predictors by accounting for model uncertainty. It is found that though stock return predictability is weak, it can still affect the investor's optimal portfolio decisions over different investment horizons.

In my PhD, I advanced the scientific exploration of the Next Generation Transit Survey (NGTS), a ground-based wide-field survey operating at ESO’s Paranal Observatory in Chile since 2016. My original contribution to knowledge is the development of novel methods to 1) estimate NGTS’ yield of planets and false positives; 2) disentangle planets from false positives; and 3) accurately characterise planets. If an exoplanet passes (transits) in front of its host star, we can measure a periodic decrease in brightness. The study of transiting exoplanets gives insight into their size, formation, bulk composition and atmospheric properties. Transit surveys are limited by their ability to identify false positives, which can mimic planets and out-number them by a hundredfold. First, I designed a novel yield simulator to optimise NGTS’ observing strategy and identification of false positives (published in Günther et al., 2017a). This showed that NGTS’ prime targets, Neptune- and Earth-sized signals, are frequently mimicked by blended eclipsing binaries, allowing me to quantify and prepare strategies for candidate vetting and follow-up. Second, I developed a centroiding algorithm for NGTS, achieving a precision of 0.25 milli-pixel in a CCD image (published in Günther et al., 2017b). With this, one can measure a shift of light during an eclipse, readily identifying unresolved blended objects. Third, I innovated a joint Bayesian fitting framework for photometry, centroids, and radial velocity cross-correlation function profiles. This allows to disentangle which object (target or blend) is causing the signal and to characterise the system. My method has already unmasked numerous false positives. Most importantly, I confirmed that a signal which was almost erroneously rejected, is in fact an exoplanet (published in Günther et al., 2018). The presented achievements minimise the contamination with blended false positives in NGTS candidates by 80%, and show a new approach for unmasking hidden exoplanets. This research enhanced the success of NGTS, and can provide guidance for future missions.

Uncertainty is pervasive in risk assessment. In ecotoxicological risk assessments, it arises from such sources as a lack of data, the simplification and abstraction of complex situations, and ambiguities in assessment endpoints (Burgman 2005; Suter 1993). When evaluating and managing risks, uncertainty needs to be explicitly considered in order to avoid erroneous decisions and to be able to make statements about the confidence that we can place in risk estimates. Although informative, previous approaches to dealing with uncertainty in ecotoxicological modelling have been found to be limited, inconsistent and often based on assumptions that may be false (Ferson & Ginzburg 1996; Suter 1998; Suter et al. 2002; van der Hoeven 2004; van Straalen 2002a; Verdonck et al. 2003a). In this thesis a Generalised Linear Modelling approach is proposed as an alternative, congruous framework for the analysis and prediction of a wide range of ecotoxicological effects. This approach was used to investigate the results of toxicity experiments on the effect of 2,4-Dichlorophenoxyacetic Acid (2,4-D) formulations and 4-Chlorophenol (4-CP, an associated breakdown product) on Daphnia carinata. Differences between frequentist Maximum Likelihood (ML) and Bayesian Markov-Chain Monte-Carlo (MCMC) approaches to statistical reasoning and model estimation were also investigated. These approaches are inferentially disparate and place different emphasis on aleatory and epistemic uncertainty (O'Hagan 2004). Bayesian MCMC and Probability Bounds Analysis methods for propagating uncertainty in risk models are also compared for the first time. For simple models, Bayesian and frequentist approaches to Generalised Linear Model (GLM) estimation were found to produce very similar results when non-informative prior distributions were used for the Bayesian models. Potency estimates and regression parameters were found to be similar for identical models, signifying that Bayesian MCMC techniques are at least a suitable and objective replacement for frequentist ML for the analysis of exposureresponse data. Applications of these techniques demonstrated that Amicide formulations of 2,4-D are more toxic to Daphnia than their unformulated, Technical Acid parent. Different results were obtained from Bayesian MCMC and ML methods when more complex models and data structures were considered. In the analysis of 4-CP toxicity, the treatment of 2 different factors as fixed or random in standard and Mixed-Effect models was found to affect variance estimates to the degree that different conclusions would be drawn from the same model, fit to the same data. Associated discrepancies in the treatment of overdispersion between ML and Bayesian MCMC analyses were also found to affect results. Bayesian MCMC techniques were found to be superior to the ML ones employed for the analysis of complex models because they enabled the correct formulation of hierarchical (nested) datastructures within a binomial logistic GLM. Application of these techniques to the analysis of results from 4-CP toxicity testing on two strains of Daphnia carinata found that between-experiment variability was greater than that within-experiments or between-strains. Perhaps surprisingly, this indicated that long-term laboratory culture had not significantly affected the sensitivity of one strain when compared to cultures of another strain that had recently been established from field populations. The results from this analysis highlighted the need for repetition of experiments, proper model formulation in complex analyses and careful consideration of the effects of pooling data on characterising variability and uncertainty. The GLM framework was used to develop three dimensional surface models of the effects of different length pulse exposures, and subsequent delayed toxicity, of 4-CP on Daphnia. These models described the relationship between exposure duration and intensity (concentration) on toxicity, and were constructed for both pulse and delayed effects. Statistical analysis of these models found that significant delayed effects occurred following the full range of pulse exposure durations, and that both exposure duration and intensity interacted significantly and concurrently with the delayed effect. These results indicated that failure to consider delayed toxicity could lead to significant underestimation of the effects of pulse exposure, and therefore increase uncertainty in risk assessments. A number of new approaches to modelling ecotoxicological risk and to propagating uncertainty were also developed and applied in this thesis. In the first of these, a method for describing and propagating uncertainty in conventional Species Sensitivity Distribution (SSD) models was described. This utilised Probability Bounds Analysis to construct a nonparametric 'probability box' on an SSD based on EC05 estimates and their confidence intervals. Predictions from this uncertain SSD and the confidence interval extrapolation methods described by Aldenberg and colleagues (2000; 2002a) were compared. It was found that the extrapolation techniques underestimated the width of uncertainty (confidence) intervals by 63% and the upper bound by 65%, when compared to the Probability Bounds (P3 Bounds) approach, which was based on actual confidence estimates derived from the original data. An alternative approach to formulating ecotoxicological risk modelling was also proposed and was based on a Binomial GLM. In this formulation, the model is first fit to the available data in order to derive mean and uncertainty estimates for the parameters. This 'uncertain' GLM model is then used to predict the risk of effect from possible or observed exposure distributions. This risk is described as a whole distribution, with a central tendency and uncertainty bounds derived from the original data and the exposure distribution (if this is also 'uncertain'). Bayesian and P-Bounds approaches to propagating uncertainty in this model were compared using an example of the risk of exposure to a hypothetical (uncertain) distribution of 4-CP for the two Daphnia strains studied. This comparison found that the Bayesian and P-Bounds approaches produced very similar mean and uncertainty estimates, with the P-bounds intervals always being wider than the Bayesian ones. This difference is due to the different methods for dealing with dependencies between model parameters by the two approaches, and is confirmation that the P-bounds approach is better suited to situations where data and knowledge are scarce. The advantages of the Bayesian risk assessment and uncertainty propagation method developed are that it allows calculation of the likelihood of any effect occurring, not just the (probability)bounds, and that the same software (WinBugs) and model construction may be used to fit regression models and predict risks simultaneously. The GLM risk modelling approaches developed here are able to explain a wide range of response shapes (including hormesis) and underlying (non-normal) distributions, and do not involve expression of the exposure-response as a probability distribution, hence solving a number of problems found with previous formulations of ecotoxicological risk. The approaches developed can also be easily extended to describe communities, include modifying factors, mixed-effects, population growth, carrying capacity and a range of other variables of interest in ecotoxicological risk assessments. While the lack of data on the toxicological effects of chemicals is the most significant source of uncertainty in ecotoxicological risk assessments today, methods such as those described here can assist by quantifying that uncertainty so that it can be communicated to stakeholders and decision makers. As new information becomes available, these techniques can be used to develop more complex models that will help to bridge the gap between the bioassay and the ecosystem.

The New Macroeconometrics may succinctly be described as the application of Bayesian analysis to the class of macroeconomic models called Dynamic Stochastic General Equilibrium (DSGE) models. A prominent local example from this research area is the development and estimation of the RAMSES model, the main macroeconomic model in use at Sveriges Riksbank.   Bayesian estimation of DSGE models is often computationally demanding. In this thesis fast algorithms for Bayesian inference are developed and tested in the context of the state space model framework implied by DSGE models. The algorithms discussed in the thesis deal with evaluation of the DSGE model likelihood function and sampling from the posterior distribution. Block Kalman filter algorithms are suggested for likelihood evaluation in large linearised DSGE models. Parallel particle filter algorithms are presented for likelihood evaluation in nonlinearly approximated DSGE models. Prefetching random walk Metropolis algorithms and adaptive hybrid sampling algorithms are suggested for posterior sampling. The generality of the algorithms, however, suggest that they should be of interest also outside the realm of macroeconometrics.

Acquises dans plusieurs centaines de bandes spectrales contiguës, les images hyperspectrales permettent d'analyser finement la composition d'une scène observée. En raison de la résolution spatiale limitée des capteurs utilisés, le spectre d'un pixel d'une image hyperspectrale résulte de la composition de plusieurs signatures associées à des matériaux distincts. À ce titre, le démélange d'images hyperspectrales vise à estimer les signatures des différents matériaux observés ainsi que leur proportion dans chacun des pixels de l'image. Pour cette analyse, il est d'usage de considérer qu'une signature spectrale unique permet de décrire un matériau donné, ce qui est généralement intrinsèque au modèle de mélange choisi. Toutefois, la signature d'un matériau présente en pratique une variabilité spectrale qui peut être significative d'une image à une autre, voire au sein d'une même image. De nombreux paramètres peuvent en être cause, tels que les conditions d'acquisitions (e.g., conditions d'illumination locales), la déclivité de la scène observée ou des interactions complexes entre la lumière incidente et les éléments observés. À défaut d'être prises en compte, ces sources de variabilité perturbent fortement les signatures extraites, tant en termes d'amplitude que de forme. De ce fait, des erreurs d'estimation peuvent apparaître, qui sont d'autant plus importantes dans le cas de procédures de démélange non-supervisées. Le but de cette thèse consiste ainsi à proposer de nouvelles méthodes de démélange pour prendre en compte efficacement ce phénomène. Nous introduisons dans un premier temps un modèle de démélange original visant à prendre explicitement en compte la variabilité spatiale des spectres purs. Les paramètres de ce modèle sont estimés à l'aide d'un algorithme d'optimisation sous contraintes. Toutefois, ce modèle s'avère sensible à la présence de variations spectrales abruptes, telles que causées par la présence de données aberrantes ou l'apparition d'un nouveau matériau lors de l'analyse d'images hyperspectrales multi-temporelles. Pour pallier ce problème, nous introduisons une procédure de démélange robuste adaptée à l'analyse d'images multi-temporelles de taille modérée. Compte tenu de la dimension importante des données étudiées, notamment dans le cas d'images multi-temporelles, nous avons par ailleurs étudié une stratégie d'estimation en ligne des différents paramètres du modèle de mélange proposé. Enfin, ce travail se conclut par l'étude d'une procédure d'estimation distribuée asynchrone, adaptée au démélange d'un grand nombre d'images hyperspectrales acquises sur une même scène à différents instants. / Acquired in hundreds of contiguous spectral bands, hyperspectral (HS) images have received an increasing interest due to the significant spectral information they convey about the materials present in a given scene. However, the limited spatial resolution of hyperspectral sensors implies that the observations are mixtures of multiple signatures corresponding to distinct materials. Hyperspectral unmixing is aimed at identifying the reference spectral signatures composing the data -- referred to as endmembers -- and their relative proportion in each pixel according to a predefined mixture model. In this context, a given material is commonly assumed to be represented by a single spectral signature. This assumption shows a first limitation, since endmembers may vary locally within a single image, or from an image to another due to varying acquisition conditions, such as declivity and possibly complex interactions between the incident light and the observed materials. Unless properly accounted for, spectral variability can have a significant impact on the shape and the amplitude of the acquired signatures, thus inducing possibly significant estimation errors during the unmixing process. A second limitation results from the significant size of HS data, which may preclude the use of batch estimation procedures commonly used in the literature, i.e., techniques exploiting all the available data at once. Such computational considerations notably become prominent to characterize endmember variability in multi-temporal HS (MTHS) images, i.e., sequences of HS images acquired over the same area at different time instants. The main objective of this thesis consists in introducing new models and unmixing procedures to account for spatial and temporal endmember variability. Endmember variability is addressed by considering an explicit variability model reminiscent of the total least squares problem, and later extended to account for time-varying signatures. The variability is first estimated using an unsupervised deterministic optimization procedure based on the Alternating Direction Method of Multipliers (ADMM). Given the sensitivity of this approach to abrupt spectral variations, a robust model formulated within a Bayesian framework is introduced. This formulation enables smooth spectral variations to be described in terms of spectral variability, and abrupt changes in terms of outliers. Finally, the computational restrictions induced by the size of the data is tackled by an online estimation algorithm. This work further investigates an asynchronous distributed estimation procedure to estimate the parameters of the proposed models.