Considered an integral part of Metropolitan France, the island of Corsica is situated nonetheless on the very periphery of the modern state that claims it. Actually situated geographically closer to Italy than to any part of France, its culture and its people are likewise more closely related to their Italians neighbors than to the rest of what Corsicans term "Continental France." Following the acquisition of Corsica, both government officials and bourgeois travelers would seek to visit the island, often recording their findings and publishing these memoirs for others to know of their travels. This concept of travel memoirs, specifically those regarding Corsica, had already been a fairly common practice among the British, as they had often placed interest in the island itself. From this group of French and British travel memoirs would come the writings of James Boswell, P. P. Pompéi, and the Baron de Beaumont, among others. Corsica becomes a place of unique setting for novels and short stories throughout the century, with tales of banditry, vendetta, and violence from the island. For those authors seeking to place their stories in Corsica, inspiration was drawn from the very travel memoirs they had read regarding the island, although often they chose to ignore them in favor of stereotypes. I have chosen three specific 19th century authors in relation to the images created by the travel memoirs of Corsica: Prosper Mérimée, Honoré de Balzac, and Guy de Maupassant. The purpose behind each author's use of the images of Corsica was very different and shows different ways that these images were used. Mérimée directly used Corsica to question the triumph of the civilized over the uncivilized, Balzac used Corsica to represent France itself, and Maupassant used Corsica to show that "reality" is really nothing more than a personal illusion. Though when publishing their travel memoirs the authors might not have expected much to come of them, they have actually influence an entire century of writers, and possibly an entire nation, with their images of Corsica.

No description available.

Operating and interacting in an environment requires the ability to manage uncertainty and to choose definite courses of action. In this thesis we look to Bayesian probability theory as the means to achieve the former, and find that through rigorous application of the rules it prescribes we can, in theory, solve problems of decision making under uncertainty. Unfortunately such methodology is intractable in realworld problems, and thus approximation of one form or another is inevitable. Many techniques make use of heuristic procedures for managing uncertainty. We note that such methods suffer unreliable performance and rely on the specification of ad-hoc variables. Performance is often judged according to long-term asymptotic performance measures which we also believe ignores the most complex and relevant parts of the problem domain. We therefore look to develop principled approximate methods that preserve the meaning of Bayesian theory but operate with the scalability of heuristics. We start doing this by looking at function approximation in continuous state and action spaces using Gaussian Processes. We develop a novel family of covariance functions which allow tractable inference methods to accommodate some of the uncertainty lost by not following full Bayesian inference. We also investigate the exploration versus exploitation tradeoff in the context of the Multi-Armed Bandit, and demonstrate that principled approximations behave close to optimal behaviour and perform significantly better than heuristics on a range of experimental test beds.

Cette thèse traite de méthodes « model-based » pour résoudre des problèmes d'apprentissage par renforcement. On considère un agent confronté à une suite de décisions et un environnement dont l'état varie selon les décisions prises par l'agent. Ce dernier reçoit tout au long de l'interaction des récompenses qui dépendent à la fois de l'action prise et de l'état de l'environnement. L'agent ne connaît pas le modèle d'interaction et a pour but de maximiser la somme des récompenses reçues à long terme. Nous considérons différents modèles d'interactions : les processus de décisions markoviens, les processus de décisions markoviens partiellement observés et les modèles de bandits. Pour ces différents modèles, nous proposons des algorithmes qui consistent à construire à chaque instant un ensemble de modèles permettant d'expliquer au mieux l'interaction entre l'agent et l'environnement. Les méthodes dites « model-based » que nous élaborons se veulent performantes tant en pratique que d'un point de vue théorique. La performance théorique des algorithmes est calculée en terme de regret qui mesure la différence entre la somme des récompenses reçues par un agent qui connaîtrait à l'avance le modèle d'interaction et celle des récompenses cumulées par l'algorithme. En particulier, ces algorithmes garantissent un bon équilibre entre l'acquisition de nouvelles connaissances sur la réaction de l'environnement (exploration) et le choix d'actions qui semblent mener à de fortes récompenses (exploitation). Nous proposons deux types de méthodes différentes pour contrôler ce compromis entre exploration et exploitation. Le premier algorithme proposé dans cette thèse consiste à suivre successivement une stratégie d'exploration, durant laquelle le modèle d'interaction est estimé, puis une stratégie d'exploitation. La durée de la phase d'exploration est contrôlée de manière adaptative ce qui permet d'obtenir un regret logarithmique dans un processus de décision markovien paramétrique même si l'état de l'environnement n'est que partiellement observé. Ce type de modèle est motivé par une application d'intérêt en radio cognitive qu'est l'accès opportuniste à un réseau de communication par un utilisateur secondaire. Les deux autres algorithmes proposés suivent des stratégies optimistes : l'agent choisit les actions optimales pour le meilleur des modèles possibles parmi l'ensemble des modèles vraisemblables. Nous construisons et analysons un tel algorithme pour un modèle de bandit paramétrique dans un cas de modèles linéaires généralisés permettant ainsi de considérer des applications telles que la gestion de publicité sur internet. Nous proposons également d'utiliser la divergence de Kullback-Leibler pour la construction de l'ensemble des modèles vraisemblables dans des algorithmes optimistes pour des processus de décision markoviens à espaces d'états et d'actions finis. L'utilisation de cette métrique améliore significativement le comportement de des algorithmes optimistes en pratique. De plus, une analyse du regret de chacun des algorithmes permet de garantir des performances théoriques similaires aux meilleurs algorithmes de l'état de l'art.

This thesis deals with a myth concept in Thomas Mann's and Ivan Olbracht's work. In regard to the tertium comparationis, which is myth, Mann's novels Joseph and His Brothers (Joseph und seine Brüder) and Doctor Faustus (Doktor Faustus) as well as Olbracht's novels The Bitter and the Sweet (Nikola Šuhaj loupežník) and Valley of Exile (Golet v údolí) are analysed. The concepts of these two authors are being compared both on a genetic level, which includes a possible influence of the tetralogy Joseph and His Brothers on Olbracht's work, and on a typological level. The second one includes analogies, which are independent of the possible direct influence. The thesis is methodological based on interpretation of fictional texts and their continuous comparison considering a literary-historical context and cultural-historical background of that time. Especially in some details, it can be considered that Thomas Mann had a direct influence on Olbracht's work in case of both of the novels. A number of analogies have been found, which exemplify a resemblance between the authors but which don't deny that Olbracht was an autonomous and creative personality at the same time. The specific myth concept of both authors harmonise with a period tendency in the modern literature.