A Hidden Markov Model generates two basic stochastic processes, a Markov chain, which is hidden, and an observation sequence. The filtering process of a Hidden Markov Model is, roughly speaking, the sequence of conditional distributions of the hidden Markov chain that is obtained as new observations are received. It is well-known, that the filtering process itself, is also a Markov chain. A classical, theoretical problem is to find conditions which implies that the distributions of the filtering process converge towards a unique limit measure. This problem goes back to a paper of D Blackwell for the case when the Markov chain takes its values in a finite set and it goes back to a paper of H Kunita for the case when the state space of the Markov chain is a compact Hausdor space. Recently, due to work by F Kochmann, J Reeds, P Chigansky and R van Handel, a necessary and sucient condition for the convergence of the distributions of the filtering process has been found for the case when the state space is finite. This condition has since been generalised to the case when the state space is denumerable. In this paper we generalise some of the previous results on convergence in distribution to the case when the Markov chain and the observation sequence of a Hidden Markov Model take their values in complete, separable, metric spaces; it has though been necessary to assume that both the transition probability function of the Markov chain and the transition probability function that generates the observation sequence have densities.

Phonetic accommodation is the process whereby speakers in an interaction modify their speech in response to their interlocutor. The social-psychological theory of Communication Accommodation Theory (Giles 1973) predicts that speakers will converge towards (become more similar to) their interlocutors in order to decrease social distance, whereas they will diverge from (become less similar to) their interlocutors to accentuate distinctiveness or show disdain. Previous studies have found that phonetic accommodation is affected by many social, situational and linguistic factors (Abrego-Collier et al. 2011; Black 2012; Babel 2009, 2010, 2012; Babel et al. 2012; Kim, Horton & Bradlow 2011; Nielsen 2011; Pardo et al. 2012). With respect to accommodation across dialects, a handful of studies have suggested that the perceptual salience of the various differences between two dialects might affect the pattern; however, these studies make conflicting predictions. Trudgill (1986) predicts that speakers will converge more towards the more salient dialectal differences, while Kim et al. (2011) and Babel (2009, 2010) suggest the opposite: that speakers will converge on the less salient differences. This thesis investigates how the perceptual salience of 6 differences between Buenos Aires Spanish and Madrid Spanish affect the pattern of phonetic accommodation in conversation. The results are considered both in terms of the magnitude of the changes that the participants make as well as the direction of the change (convergence or divergence). The results show that perceptual salience has a significant effect on the magnitude of the change, with all participants making greater changes as perceptual salience increases. On the other hand, perceptual salience was found not to have a consistent effect for all speakers on the likelihood of converging or diverging on the dialectal differences. I argue that the lack of consistent effect of salience on the direction of the change stems from individual differences in motivation to take on the opposing dialect norms and issues of personal identity, whereas the very consistent effect of salience on the magnitude of the change suggests that there is something more basic or systematic about how salience interacts with the extent to which speakers accommodate.

In this thesis, studies associated with the stability analyses of the box-shaped dump-site culvert constructed in Tinaz Surface Mine of Turkish Coal Enterprises (TKi) are presented. In addition, stability conditions of other culvert alternatives are evaluated. Existence of creeks in a surface mining area is a significant factor to be considered in selection of dump-site location. Since, the dumped overburden material on the valley acts as a barrier and behaves like a dam causing flood problem behind the dump-site. TKi engineers prevented the flood potential that might have occurred behind the dump-site by constructing a 480-meter long reinforced-concrete culvert on the downstream of Gevenez Creek Valley. However, considerable amount of deformations occurred in the first 100 meters of the culvert, as a result of overburden material being replaced on this structure. In order to determine the failure mechanism associated with the culvert, a series of numerical modeling analyses were carried out utilizing back analysis technique. The validity of the numerical model was justified by convergence measurements and observations carried out inside the culvert as overburden material being replaced on the stable part of this structure. Finally, based on the numerical model developed, the stability of other culvert alternatives that could be used in future projects were evaluated considering different embankment conditions (positive projecting and negative projecting), bedding conditions (impermissible, ordinary, first-class and concrete cradle), culvert shapes (box and circular) and dumping conditions.

The purpose of this paper is to discuss the differences between accounting for fixed assets under IFRS and U.S. GAAP. Primarily the discussion will be driven by the question: which standard should the FASB lobby for in its joint effort with the IASB to converge U.S. GAAP with IFRS? The paper will start by establishing that financial reporting, as it has evolved in the United States, was developed primarily to assist in the accountability relationships between management and both credit and equity investors. From there the paper will look at which standard is better suited to fulfill this goal and enhance financial reporting. It will first discuss the differences of historical cost accounting versus market values in the income statement. Next, the paper will look at the differences created in the balance sheet by the two methods. Finally, the paper will consider additional implications of switching to a market value standard. This paper will attempt to prove that historical cost accounting is the more reliable method of the two, provides an equally if not more relevant income measure, and that market values do not enhance the financial statements in any clear way. Particularly in any fashion that justifies the additional costs to the firm and to investors in debt and equity that market values would create.

Dans ce mémoire, je développe un modèle d'équilibre général multirégional avec présence d'une infrastructure locale et la possibilité pour les ménages de migrer d'une région à une autre. Lorsque l'infrastructure est publique et productive, celle-ci est davantage présente dans les régions les plus peuplées, générant ainsi des écarts régionaux. Plus une région est peuplée et plus elle peut investir en infrastructure, améliorant ainsi sa productivité, le niveau des salaires et donc son attractivité. Toutefois, des externalités de congestion peuvent apparaître lorsque la région devient trop peuplée, réduisant ainsi l'utilité du ménage. À partir de ce cadre analytique, il est alors possible d'étudier le comportement des ménages et la dynamique à long terme des régions. Le modèle est étalonné à des données concernant les 96 départements de la France métropolitaine. Le modèle est alors capable de reproduire le paradoxe régional observé : les ménages migrent des départements les plus riches vers les départements les plus pauvres. Dans ce cas précis, le sens des migrations est essentiellement expliqué par les effets de congestions qui dominent les effets de richesse, conduisant ainsi les ménages à préférer des départements moins peuplés et donc moins riches. Le modèle est ensuite simulé afin d'en évaluer la dynamique à long terme. Je montre qu'il existe un mécanisme naturel d'ajustement régional conduisant à la convergence des principaux agrégats de l'ensemble des départements. Toutefois, dans les réalités les ménages sont soumis à de fortes rigidités à la mobilité ce qui ralentit considérablement le processus de convergence régionale. Le temps d'ajustement est évalué à plusieurs siècles. Les cas de quelques départements représentatifs sont analysés à la fin de ce mémoire. Il est montré que l'ensemble des départements ne convergent pas de la même manière. D'une part, je montre que des départements fortement peuplés comme la Seine-et-Marne ou faiblement peuplés comme le Loir-et-Cher convergent de manière monotone. D'autre part, je montre que des départements de taille moyenne, tel que l'Indre-et-Loire, convergent dans un premier temps, puis divergent dans un second temps, pour ensuite converger, présentant ainsi une tendance oscillatoire à long terme. ______________________________________________________________________________ MOTS-CLÉS DE L’AUTEUR : congestion, convergence régionale, infrastructure, migration.

We study the zeroth-, first-, and second-order algorithms proposed by Tuncel. The zeroth-order algorithms are the generalization of the classic primal-dual affine-scaling methods, and have a strong connection with the quasi-Newton method. Although the zeroth-order algorithms have the property of strict monotone decrease in both primal and dual objective values, they may not converge. We give an illustrative example as well as an algebraic proof to show that the zeroth-order algorithms do not converge to an optimal solution in some cases. The second-order algorithms use the gradients and Hessians of the barrier functions. Tuncel has shown that all second-order algorithms have a polynomial iteration bound. The second-order algorithms have a range of primal-dual scaling matrices to be chosen. We give a method to construct such a primal-dual scaling matrix. We then analyze a new centrality measure. This centrality measure appeared in both first- and second-order algorithms. We compare the neighbourhood defined by this centrality measure with other known neighbourhoods. We then analyze how this centrality measure changes in the next iteration in terms of the step length and some other information of the current iteration.

George Klosko attempts to solve a problem put forth by Rawls, namely how to create a persisting, just and stable liberal democracy in light of pluralism. He believes Rawls has failed at this task through the employment of political constructivism. Klosko claims that since Rawls does not utilize actual views within the existing public to form principles of justice, his method would fail to reach an overlapping consensus. As an alternative, Klosko proposes the method of convergence, which utilizes actual societal views to find overlapping concepts that inform the principles of justice. My argument is that Klosko misconstrues the method and aims of political constructivism. Klosko seems to incorrectly believe that stability is primary to establishing a liberal democracy, whereas it is secondary to the achievement of justice. Because of this error, Klosko’s method of convergence potentially has the consequence of creating a society which is stable but unjust.

This thesis explores ways evangelical Christian communicators remediate traditional ministry functions and community formation onto new media platforms. This exploration is framed by a discussion of literacy and digital composing reflecting Stuart Selber’s multiliteracy approach to teaching digital composition. The author positions evangelical churches’ approaches to texts, community, education, and communication as components of a distinct literacy that is often at odds with values, controls, and cultures found on the Internet and in new media. Discovering how church communicators use new media, how their education prepares them for effective digital communication, and how external sources, such as expert authors, aid the transition from print to new media helps us understand the gap between Selber’s ideal multiliteracy and the reality of new media literacy for this group. This also expands our understanding of digital composition, and the role it plays in both the classroom and in all students’ greater lives.

As the capabilities of autonomous vehicles increase, their use in situations that are dangerous or dull for humans is becoming more popular. Autonomous systems are currently being used in several military and civilian domains, including search and rescue operations, disaster relief coordination, infrastructure inspection and surveillance missions. In order to perform high level mission autonomy tasks, a method is required for the vehicle to localize itself, as well as generate a map of the environment. Algorithms which allow the vehicle to concurrently localize and create a map of its surroundings are known as solutions to the Simultaneous Localization and Mapping (SLAM) problem. Certain high level tasks, such as drivability analysis and obstacle avoidance, benefit from the use of a dense map of the environment, and are typically generated with the use of point cloud data. The point cloud data is incorporated into SLAM algorithms with scan registration techniques, which determine the relative transformation between two sufficiently overlapping point clouds. The Normal Distributions Transform (NDT) algorithm is a promising method for scan registration, however many issues with the NDT approach exist, including a poor convergence basin, discontinuities in the NDT cost function, and unreliable pose estimation in sparse, outdoor environments. This thesis presents methods to overcome the shortcomings of the NDT algorithm, in both 2D and 3D scenarios. To improve the convergence basin of NDT for 2D scan registration, the Multi-Scale k-Means NDT (MSKM-NDT) algorithm is presented, which divides a 2D point cloud using k-means clustering and performs the scan registration optimization over multiple scales of clustering. The k-means clustering approach generates fewer Gaussian distributions when compared to the standard NDT algorithm, allowing for evaluation of the cost function across all Gaussian clusters. Cost evaluation across all the clusters guarantees that the optimization will converge, as it resolves the issue of discontinuities in the cost function found in the standard NDT algorithm. Experiments demonstrate that the MSKM-NDT approach can be used to register partially overlapping scans with large initial transformation error, and that the convergence basin of MSKM-NDT is superior to NDT for the same test data. As k-means clustering does not scale well to 3D, the Segmented Greedy Cluster NDT (SGC-NDT) method is proposed as an alternative approach to improve and guarantee convergence using 3D point clouds that contain points corresponding to the ground of the environment. The SGC-NDT algorithm segments the ground points using a Gaussian Process (GP) regression model and performs clustering of the non ground points using a greedy method. The greedy clustering extracts natural features in the environment and generates Gaussian clusters to be used within the NDT framework for scan registration. Segmentation of the ground plane and generation of the Gaussian distributions using natural features results in fewer Gaussian distributions when compared to the standard NDT algorithm. Similar to MSKM-NDT, the cost function can be evaluated across all the clusters in the scan, resulting in a smooth and continuous cost function that guarantees convergence of the optimization. Experiments demonstrate that the SGC-NDT algorithm results in scan registrations with higher accuracy and better convergence properties than other state-of-the-art methods for both urban and forested environments.

We give an upper bound for the norm distance of (0,1) -valued Markov-exchangeable random variables to mixtures of distributions of Markov processes. A Markov-exchangeable random variable has a distribution that depends only on the starting value and the number of transitions 0-0, 0-1, 1-0 and 1-1. We show that if, for increasing length of variables, the norm distance to mixtures of Markov processes goes to 0, the rate of this convergence may be arbitrarily slow. (author's abstract) / Series: Forschungsberichte / Institut für Statistik