none

<p>Given the question of how to further decrease childhood mortality and attain the fourth MDG in Bangladesh, the determinants of childhood mortality and successful health intervention programmes in a rural area of Bangladesh are examined in this paper. The binominal logit regression analysis, on Matlab HDSS data from 2001 to 2005, indicates that the child’s birth order, outcome of mother’s previous pregnancy, mother’s age, mother’s education, economic condition of the household, immunization, and place of delivery are important determining factors of childhood mortality. Interview discussions show that the delivery of health services is a determining factor for successful health intervention programmes. It is worth to note that childhood mortality levels are no longer significantly lower in the treatment area of Matlab. Furthermore, the intervention programmes in the area require continuous reform, especially in the fields of birth assistance and injury prevention.</p>

Given the question of how to further decrease childhood mortality and attain the fourth MDG in Bangladesh, the determinants of childhood mortality and successful health intervention programmes in a rural area of Bangladesh are examined in this paper. The binominal logit regression analysis, on Matlab HDSS data from 2001 to 2005, indicates that the child’s birth order, outcome of mother’s previous pregnancy, mother’s age, mother’s education, economic condition of the household, immunization, and place of delivery are important determining factors of childhood mortality. Interview discussions show that the delivery of health services is a determining factor for successful health intervention programmes. It is worth to note that childhood mortality levels are no longer significantly lower in the treatment area of Matlab. Furthermore, the intervention programmes in the area require continuous reform, especially in the fields of birth assistance and injury prevention.

The challenges posed by climate change are a threat to human well-being as well as to natural ecosystems, but researchers indicate that awareness of and concern about climate change varies considerably. In this thesis, we investigate the socio-demographic factors associated with the perception of climate change, analysing four data sets collected from four surveys on urban middle-income households in Mexico. Our empirical strategy relies on the estimation of logit regressions on a binary variable defining whether respondents consider climate change as an important issue. Results indicate that age, gender, education, employment status, household size, and having in the household a member with a respiratory illness are important predictors that shape Mexicans’ perception of how important climate change is. It is important to know the determinants that effect climate change perception in order to develop sustainable policies to mitigate the risks of climate change.

一般而言， 銀行決定貸放與否，除了取決於客戶所提供之擔保品、保證 人之財力大小外， 徵信工作人員憑著自己過去經驗主觀之判定亦是一重 要之影響因素。 然而人為所牽涉之因素過於複雜且多變，若稍不謹慎極 可能犯下極大的錯失。 也因此銀行必須發展一套客觀之信用評估模型來 分析客戶的信用，以求降低放款的風險。而一般國內之銀行信用評估模型 及理論大多以混和行業別為研究對象，且多以財務比率分析為主。 但不 同的產業其財務特性有所差異，因此本研究便試圖以塑膠業建立單一產業 之信用評估模型， 以尋求建立更適合之信用評估模型。本研究首先根據 國內外文獻中整理出可供研究之若干財務變數， 再參酌目前銀行所使用 作為徵信評估之財務變數， 由於本模式是採本國塑膠業之資料， 且應用 對象亦主要針對本國銀行，故決定採用目前省屬行庫做徵信評估時所採行 的 17 個財務變數做分析。而後再以塑膠加工業為研究對象， 蒐集民國 79 年至 81 年三十家違約公司與五十七家履約公司共八十七家公司的財 務報表資料以計算其財務比率， 接著便計算此 17 個財務變數是否符合 常態分配之假設，以決定後續所採行之統計方法應為何。 另外根據因素 分析與不採用因素分析分別進行變數之選擇及相關性之分析，以尋找具有 解釋能力之財務變數。 最後以上述所得到之財務變數做自變數， 並將樣 本區分為原始樣本、預測樣本以及全體樣本兩群分別建立 Logit 迴歸模 型，以得到不同分析方法之總正確率、型一誤差、型二誤差及加權效率性 。 爾後再根據結果做較佳模式之選擇與評估，並加以解釋。根據實證結 果顯示， 依不同樣本及不同方法所建立的四個放款信用評估模型中， 經 比較其總正確率及加權效率性的高低後，採用因素分析的模型將優於不採 用因素分析的模型， 而全體樣本所建立的模型會優於原始樣本所建立的 模型。 而由此所得到最佳的信用評估模型為以全體樣本採用因素分析所 得的模型。另外， 本研究是以塑膠業單一產業為研究對象來建立信用評 估模型，與過去採混和產業別的分析有所不同。 而與過去的分析相較後 ，得知其正確區別效果確實較混和產業別所建立的模型為佳。

This thesis deals with the development, implementation and application of statistical modeling techniques which can be employed in the analysis of credit ratings. Credit ratings are one of the most widely used measures of credit risk and are relevant for a wide array of financial market participants, from investors, as part of their investment decision process, to regulators and legislators as a means of measuring and limiting risk. The majority of credit ratings is produced by the "Big Three" credit rating agencies Standard & Poors', Moody's and Fitch. Especially in the light of the 2007-2009 financial crisis, these rating agencies have been strongly criticized for failing to assess risk accurately and for the lack of transparency in their rating methodology. However, they continue to maintain a powerful role as financial market participants and have a huge impact on the cost of funding. These points of criticism call for the development of modeling techniques that can 1) facilitate an understanding of the factors that drive the rating agencies' evaluations, 2) generate insights into the rating patterns that these agencies exhibit. This dissertation consists of three research articles. The first one focuses on variable selection and assessment of variable importance in accounting-based models of credit risk. The credit risk measure employed in the study is derived from credit ratings assigned by ratings agencies Standard & Poors' and Moody's. To deal with the lack of theoretical foundation specific to this type of models, state-of-the-art statistical methods are employed. Different models are compared based on a predictive criterion and model uncertainty is accounted for in a Bayesian setting. Parsimonious models are identified after applying the proposed techniques. The second paper proposes the class of multivariate ordinal regression models for the modeling of credit ratings. The model class is motivated by the fact that correlated ordinal data arises naturally in the context of credit ratings. From a methodological point of view, we extend existing model specifications in several directions by allowing, among others, for a flexible covariate dependent correlation structure between the continuous variables underlying the ordinal credit ratings. The estimation of the proposed models is performed using composite likelihood methods. Insights into the heterogeneity among the "Big Three" are gained when applying this model class to the multiple credit ratings dataset. A comprehensive simulation study on the performance of the estimators is provided. The third research paper deals with the implementation and application of the model class introduced in the second article. In order to make the class of multivariate ordinal regression models more accessible, the R package mvord and the complementary paper included in this dissertation have been developed. The mvord package is available on the "Comprehensive R Archive Network" (CRAN) for free download and enhances the available ready-to-use statistical software for the analysis of correlated ordinal data. In the creation of the package a strong emphasis has been put on developing a user-friendly and flexible design. The user-friendly design allows end users to estimate in an easy way sophisticated models from the implemented model class. The end users the package appeals to are practitioners and researchers who deal with correlated ordinal data in various areas of application, ranging from credit risk to medicine or psychology.

Nombreuses sont les villes françaises à abriter des éléments d’architecture militaire (citadelles, casernes, bastions,murailles, etc.). Bien que les fonctions défensives à l’origine de ces ouvrages soient peu à peu tombées en désuétude,ceux-ci sont ancrés dans les tissus urbains des cités qui les abritent aujourd’hui encore. Ils participent des morphologiesurbaines, des schémas fonctionnels et, plus globalement, de la manière dont la notion de ville est appréhendée.Les villes font dorénavant face à de nouveaux enjeux qui poussent à renouveler les cadres au travers desquelles ellessont traditionnellement approchées. L’étalement urbain, ses conséquences, la nécessité de tendre vers un urbanismerenouvelé et plus durable amènent à porter un nouveau regard sur les fortifications urbaines. Au-delà de son aspectsymbolique, ce type d’architecture est marqué par des caractéristiques spatiales hors-normes (localisation, volumétrie,géométrie) qui engagent à l’analyser au travers des prismes méthodologiques et conceptuels que fournit la géographie.Comment l’héritage des fortifications urbaines peut-il aujourd’hui être mobilisé afin de créer un espace urbain cohérent,tant sur le plan spatial que social ou symbolique et ainsi répondre aux aspiration des villes actuelles ?La première partie de la thèse porte sur les enjeux auxquels les “villes fortes” sont aujourd’hui confrontées (modernisation,croissance étalement urbain, etc.) et la manière dont leurs “ouvrages lourds” sont affectés. La deuxième partiedu travail s’attache à caractériser le phénomène des “villes fortes” en France aujourd’hui par le biais d’approchesdescriptives, empiriques et statistiques qui permettent de formaliser et d’analyser les logiques et les règles qui régissentle devenir des anciens ouvrages défensifs. La dernière partie de la thèse étudie le potentiel d’urbanisation des zones etouvrages militaires et propose de mesurer l’impact de leur transformation en logements dans le cadre d’un scénariode renouvellement urbain compact pour la ville de Besançon. Il s’agit d’évaluer les conséquences de telles mesuressur le schéma fonctionnel des villes et les formes urbaines. L’ensemble de ces travaux montre l’intérêt du recours auxouvrages défensifs dans le cadre d’un urbanisme à la fois renouvelé, mais aussi plus “durable”, en conformité avec lesformes urbaines, les demandes contemporaines et les projets des villes. / A large number of french cities host military historical edifices (citadels, barracks, bastions, defensive walls, etc.). Although their initial defensive functions have been lost over time, these edifices remain deeply rooted in the urban fabric of their host cities. They continue exerting an impact on these cities’ urban morphology and modern-time functions as well as the way in which the concept of city is understood. Cities nowadays face some new challenges,the increasing awareness of urban sprawl and its consequences, coupled with an urge to promote a renewed and sustainable urbanism, invites us to adopt new approaches to study urban fortifications. In addition to their symbolic aspect, fortifications are characterized by their out-of-the-common spatial measure (location, volume and geometry),requiring researchers to use methodologies and geographical concepts for their academic endeavour. Against this background : how can urban fortifications be used to create urban space that is spatially, socially and symbolically coherent with their historical heritage and that also satisfies the functional need of modern cities ?The first part of this doctoral dissertation present modern-day challenges that fortified cities face (modernisation,urban expansion and sprawl, etc.), and the ways in which their “heavy buildings” are affected. The second part sheds light on the relationships between cities and their fortifications, thanks to descriptive, empirical and statistical analyses. The statistical study noticeably builds on a database to examine the rationale and rules underpinning there habilitation of ancient military edifices. The last part of the dissertation proposes an evaluation of the potential of rehabilitating ancient military edifices and spaces for modern urban usage. For this purpose the impact of transforming these fortifications into residential blocks is assessed in a fortified French city (Besançon). This assessment aims at evaluating the consequences of such urban planning measures on urban morphology and functions. Overall, the analysis provided in that doctoral dissertation demonstrates that ancient military buildings have their place in contemporary urban planning. They help to achieve a renewed and sustainable urban design in line with the urban morphology,modern social and functional requirements and the development perspectives in their host cities.