區域間由於發展背景與功能取向不同，以致各地區社會福利支出比例有所差異，對治安與生活品質亦造成不同的影響。社會福利制度是促進整體社會平等、維持人民生存的重要體制；犯罪人口的多寡，則是影響國家成長與進步的重要社會問題之一。本研究主要目的為探討台灣地區社會福利支出對總犯罪率、各年齡層（少年、青年與成年）及失業者犯罪人口率之影響程度，並進一步分析不同社會福利補助支出對總體犯罪率及上述不同群體的個別影響方向。研究範圍為1999-2009年台灣地區23個縣市別的追蹤資料（panel data），利用雙因子固定效果模型（two-way fixed-effect model）同時分析各縣市之個別特質與時間趨勢效果，可對討論議題提供精確的估計結果。研究結果顯示，台灣各縣市社會福利支出與總體犯罪率間不存有顯著關係。進一步分析五種社會福利補助支出，發現家庭生活扶助對於改善整體及各年齡層的犯罪問題皆具有良好效果；其他如就學生活補助、醫療補助與急難救助對部分族群之犯罪問題亦有正面幫助。唯失業給付對犯罪問題並無正面影響。因此建議可藉由調整社會福利支出的比例以降低犯罪問題的發生。 / As the mode of growth and orientation policy in various area are different, resulting in proportion of social welfare spending and influence on public security and quality of life are also different. This study aims to investigate the influence of social welfare spending as well as different social welfare assistance on the total crime rate and crime rate of population of specific age groups and unemployed in Taiwan. Using the cross-regional panel data of 23 counties during the period 1999-2009 in Taiwan. The empirical results from two-way fixed-effect model found that there is no significant relationship between social welfare spending and crime rate. However, analyzing the further five social welfare assistances with regard to improve crime problem has been proved, especially living support assistance. Therefore, this study suggests that the criminal coping strategy is proposed to adjust proportion social welfare spending to reduce criminal problems.

中國大陸自改革開放之後，其經濟成長之快速，使得全球眾多的學者投入中國大陸議題的研究。隨著世界的開發，有限的資源不斷被使用，造成資源愈來愈少，也由於資源的稀少性，因此市場經濟方為競爭的經濟。參與競爭必須首先具備競爭力，作為國民經濟重要組成單元的區域經濟，如何應對日趨激烈的競爭，如何培育構建、提升保持具有區域特色的競爭力，就成為迫在眉睫的問題 本研究之研究目的主要有以下幾點：第一，希望能藉由相關理論文獻，解釋財政地方分權對中國省、市之間競爭力是否會有所影響或是關聯，將以此為本研究之理論基礎，並且進行實證的檢驗；第二，將以1994年中國財政改革以後之財政相關資料，利用中國31個省、市的追蹤資料，資料蒐集期間涵蓋2000年至2008年，以各年各省、市的綜合競爭力作為衡量該省、市的競爭力指標，以期能夠了解各地區競爭力的影響情形； 第三，本研究建立一個二階段固定效果模型，來檢視中國31個省市財政分權對其綜合競爭力的影響，並將各省市的情形做歸納；第四，從實證模型中發現，財政地方分權與區域綜合競爭力間的關聯性為一非線性關係且具有U型曲線關係；最後，利用實證結果來提供具體的政策建議。

城市化意味着城市的經濟結構、社會結構和空間結構變遷。隨着中國大陸城市化的推進，城市人口遷移、建成區面積增長，城市發展質量逐步提高，而同期內城市商品住宅價格亦呈現普遍而明顯的上漲趨勢。 過去雖有許多國內外文獻對全國城市商品住宅價格的影響因素進行了探討，但少有文獻從城市化的層面着眼進行研究；在部分集中討論城市化效應的文獻中，亦存在以省域代替城市作爲研究對象，以及採用橫截面數據進行實證研究，未能體現不同城市化和社會經濟發展階段之變動趨勢的缺憾。 有鑑於此，本文使用中國大陸2003年至2011年不同規模城市的相關追蹤資料，以“人口城市化”“土地城市化”和城市發展質量優化作爲城市化的主要體現，用從業人口、建設用地面積、房地產開發投資額等一系列指標作爲自變數構建追蹤資料固定效應模型，分析城市化對不同規模城市商品住宅價格的影響，探究中國大陸社會經濟重要發展來源的城市化能否繼續維持城市商品住宅市場的穩定。實證結果顯示，城市化發展對城市商品住宅價格上漲的貢獻明顯，而人口城市化的作用較之土地城市化更爲顯著；大城市和中小城市在城市化發展過程中商品住宅價格影響因素亦有所不同，城市發展中由政府主導的城市土地開發利用與基礎設施建設應集約化發展，土地城市化必須與人口城市化以及城市發展質量提升相互協調。 / Urbanization means changes of cities ‘economic structure, social structure and spatial structure. With the development of Chinese urbanization, the urban population and construction land area is growing, the urban development quality is improving gradually, in the same period city residential housing prices also presents common and obvious rising trend. In the past, although there were many domestic and external literature focusing on the influence factors of the national urban residential housing prices, but there were few literature based the study on the impact of urbanization; many of literatures in which the urbanization effect discussed, also taken provincial data as the research object instead of city , using cross-sectional data for empirical research, so failed to reflect changes in different stages of urbanization and social and economic development trend. In light of this, this article refer to mainland China from 2003 to 2011, 213 level panel data, related to "land urbanization" and "population urbanization" quality optimization as the main embodiment of urbanization, taking urban working population, construction land area, real estate investment and a series of indicators as independent variables to construct panel fixed effects model, in order to analyze the effect of urbanization on the urban residential property prices. We expect to explore whether if mainland China's urbanization, which is a source of important social and economic development of the nation, will continue to maintain the stability of the urban residential market. Empirical results indicates that the influence of urbanization on city residential housing price is evident, besides population urbanization plays a more important role than land urbanization. In the process of urbanization, the influential factors of residential housing price varying in metropolis and small cities, therefore development of urban land dominated by local government ought to realize intensification, and land urbanization should be coordinated with population urbanization and city development quality.

The distribution of mountain permafrost is generally modeled using a combination of statistical techniques and empirical variables. Such models, based on topographic, climatic and geomorphological predictors of permafrost, have been widely used to estimate the spatial distribution of mountain permafrost in North America and Europe. However at present, little knowledge about the distribution and characteristics of mountain permafrost is available for the Andes. In addition, the effects of climate change on slope stability and the hydrological system, and the pressure of mining activities have increased concerns about the knowledge of mountain permafrost in the Andes. In order to model permafrost distribution in the semi-arid Chilean Andes between ~29°S and 32°S, an inventory of rock glaciers is carried out to obtain a variable indicative of the presence and absence of permafrost conditions. Then a Linear Mixed-Effects Model (LMEM) is used to determine the spatial distribution of Mean Annual Air Temperature (MAATs), which is then used as one of the predictors of permafrost occurrence. Later, a Generalized Additive Model (GAM) with a logistic link function is used to predict permafrost occurrence in debris surfaces within the study area. Within the study area, 3575 rock glaciers were inventoried. Of these, 1075 were classified as active, 493 as inactive, 343 as intact and 1664 as relict forms, based on visual interpretation of satellite imagery. Many of the rock glaciers (~60-80%) are situated at positive MAAT, and the number of rock glaciers at negative MAAT greatly decreases from north to south. The results of spatial temperature distribution modeling indicated that the temperature changes by -0.71°C per each 100 m increase in altitude, and that there is a 4°C temperature difference between the northern and southern part of the study area. The altitudinal position of the 0°C MAAT isotherm is situated at ~4250 m a.s.l. in the northern (29°S) section and drops latitudinally to ~4000 m a.s.l. in the southern section (32°S) of the study area. For permafrost modeling purposes, 1911 rock glaciers (active, inactive and intact forms) were categorized into the class indicative of permafrost presence and 1664 (relict forms) as non-permafrost. The predictors MAAT and Potential Incoming Solar Radiation (PISR) and their nonlinear interaction were modeled by the GAM using LOESS smoothing function. A temperature offset term was applied to reduce the overestimation of permafrost occurrence in debris surface areas due to the use of rock glaciers as permafrost proxies. The dependency between the predictor variables shows that a high amount of PISR has a greater effect at positive MAAT levels than in negative ones. The GAM for permafrost distribution achieved an acceptable discrimination capability between permafrost classes (area under the ROC curve ~0.76). Considering a permafrost probability score (PPS) ≥ 0.5 and excluding steep bedrock and glacier surfaces, mountain permafrost can be potentially present in up to about 6.8% (2636 km2) of the study area, whereas with a PPS ≥ 0.75, the potential permafrost area decreases to 2.7% (1051 km2). Areas with the highest PPS are spatially concentrated in the north section of the study area where altitude rises considerably (the Huasco and Elqui watersheds), while permafrost is almost absent in the southern section where the topography is considerably lower (Limarí and Choapa watersheds). This research shows that the potential mountain permafrost distribution can be spatially modeled using topoclimatic information and rock glacier inventories. Furthermore, the results have provided the first local estimation of permafrost distribution in the semi-arid Chilean Andes. The results obtained can be used for local environmental planning and to aid future research in periglacial topics.

中國大陸自改革開放之後，其經濟成長之快速，成功歸究於自由的市場經濟。然而2008金融海嘯爆發後，使得各國開始檢視過度的分權和市場自由化的適度性。中央政府和地方政府，如何在權利和稅收的分配上，達成一最適的規模？如何有效運用財政地方分權？將是一門複雜且重要的議題。 本研究之研究目的主要有以下幾點：第一，希望能藉由相關理論文獻，解釋財政地方分權對中國省、市之間地方財政赤字是否會有所影響或是關聯，將以此為本研究之理論基礎，並且進行實證的檢驗；第二，將以1994年以後中國財政改革以後之財政相關資料，利用中國31個省、市的追蹤資料，資料蒐集期間涵蓋1995年至2010年，以各年各省、市的財政赤字作為衡量該省、市的財政情形，以期能夠了解各地區財政的影響情形； 第三，本研究建立一個二因子固定效果模型，來檢視中國31個省市財政分權對其地方財政赤字的影響，並將各省市的情形做歸納；第四，從實證模型中發現，財政地方分權與地方財政赤字間的關聯性為一非線性關係且具有U型曲線關係；最後，利用實證結果來提供具體的政策建議。 / Fiscal decentralization is considered as one of the successful institutional reforms to promote the development of China.In order to attract the resource, the regions have to improve or maintain their finance. Therefore, what issues will improve the provincial local government deficit becoming an immense problem. How to use fiscal decentralization tool？It will be an important issue. The research purpose of this literature is to use the empirical model with the panel data which includes 31 provinces, cities and regions in China during the period of 1995 to 2010 as well as to search the following question. First , analyse the provincial local government deficit and find out the determinative factors of regional deficit. Second, in order to realize the precise relationship between the fiscal decentralization and the provincial local government deficit, this study establishes 2 ways fixed effect model. Finally, to reducing the mistakes occurred in positive models and enabling the study more rigorous, this study uses more methods to test the models and the result. Finally, to reducing the mistakes occurred in positive models and enabling the study more rigorous, this study uses more methods to test the models and the result.

This thesis consists of two parts. The first part focuses on design and analysis for computer experiments and the second part deals with binary time series and its application to kinetic studies in micropipette experiments. The first part of the thesis addresses three problems. The first problem is concerned with optimal design of computer experiments. Latin hypercube designs (LHDs) have been used extensively for computer experiments. A multi-objective optimization approach is proposed to find good LHDs by combining correlation and distance performance measures. Several examples are presented to show that the obtained designs are good in terms of both criteria. The second problem is related to the analysis of computer experiments. Kriging is the most popular method for approximating complex computer models. Here a modified kriging method is proposed, which has an unknown mean model. Therefore it is called blind kriging. The unknown mean model is identified from experimental data using a Bayesian variable selection technique. Many examples are presented which show remarkable improvement in prediction using blind kriging over ordinary kriging. The third problem is related to computer experiments with nested and branching factors. Design and analysis of experiments with branching and nested factors are challenging and have not received much attention in the literature. Motivated by a computer experiment in a machining process, we develop optimal LHDs and kriging methods that can accommodate branching and nested factors. Through the application of the proposed methods, optimal machining conditions and tool edge geometry are attained, which resulted in a remarkable improvement in the machining process. The second part of the thesis deals with binary time series analysis with application to cell adhesion frequency experiments. Motivated by the analysis of repeated adhesion tests, a binary time series model incorporating random effects is developed in this chapter. A goodness-of-fit statistic is introduced to assess the adequacy of distribution assumptions on the dependent binary data with random effects. Application of the proposed methodology to real data from a T-cell experiment reveals some interesting information. These results provide some quantitative evidence to the speculation that cells can have ¡§memory¡¨ in their adhesion behavior.

As linhas de distribuição de energia estão freqüentemente expostas a sobretensões causadas por descargas atmosféricas diretas e indiretas. As formas de onda dessas sobretensões têm uma faixa de variação muito ampla e podem diferir bastante do impulso atmosférico normalizado utilizado em ensaios para verificação da adequação dos projetos das isolações dos equipamentos frente a sobretensões atmosféricas (1,2 / 50 microssegundos). É fato conhecido que a suportabilidade das isolações depende não só da amplitude como da forma de onda das solicitações. Diferentes modelos têm sido propostos para se estimar o desempenho das isolações frente a impulsos não normalizados, sendo o modelo do efeito disruptivo (disruptive effect model) um dos mais utilizados. Existem, contudo, diferentes métodos de aplicação desse modelo, ou seja, diferentes formas de se estimar os parâmetros necessários para a sua aplicação. Este trabalho visa avaliar o comportamento dielétrico de isoladores de média tensão e analisar os principais métodos para estimativa da suportabilidade desses equipamentos frente a sobretensões atmosféricas com formas de onda diferentes da normalizada. Para essa avaliação foram realizados ensaios em um isolador tipo pino, de porcelana, com tensão nominal de 15 kV, nos quais foram utilizadas, além do impulso atmosférico normalizado, outras ondas selecionadas com base em resultados de medição e de cálculo. Modificações realizadas no circuito de um gerador de impulsos de alta tensão convencional permitiram a geração de tensões com formas de onda bastante semelhantes às de sobretensões induzidas por descargas atmosféricas tanto em linhas de tamanho natural como em experimentos realizados com modelo reduzido. São apresentados e discutidos os resultados dos ensaios de tensão disruptiva de impulso atmosférico a 50 % e as curvas tensão-tempo (U x t) obtidas para cada impulso, considerando ambas as polaridades. A avaliação dos métodos de aplicação do modelo do efeito disruptivo foi realizada com base em comparações entre as curvas tensão-tempo obtidas nos ensaios e as curvas previstas por cada modelo, para cada uma das ondas selecionadas. / Overhead distribution lines are often exposed to lightning overvoltages, whose waveshapes vary widely and can differ substantially from the standard impulse voltage waveshape used to test electric equipment insulation against lightning surges (1.2 / 50 microseconds wave). It is well known that the voltage withstand capability of insulation depends not only on the amplitude but also on the voltage waveshape. Different models have been proposed for predicting the strength of insulation subjected to impulses of non-standard waveshapes. One of the most commonly used is the \"disruptive effect model\". There are, however, different methods of applying this model, that is, different ways of estimating the parameters needed for its application. This thesis aims at evaluating the dielectric behavior of medium voltage insulators subjected to impulses of non-standard waveshapes, as well as at evaluating the main methods for predicting their dielectric strength against such impulses. For the analysis, tests were performed on a pin type porcelain insulator with rated voltage of 15 kV, using, besides the standard lightning impulse voltage waveshape, other s waveshapes selected based on the characteristics of measured and calculated lightning overvoltages. Modifications made to the circuit of a conventional impulse voltage generator allowed to obtain voltage waveshapes very similar to those of lightning-induced voltages measured in experiments conducted both in lines of natural size and in reduced model. The test results relative to the critical lightning impulse flashover voltage (U50) and the volt-time characteristics obtained for the positive and negative polarities of each waveshape are presented and discussed. The evaluation of the methods of determining the parameters of the disruptive effect model was based on comparisons between the volt-time curves obtained from the laboratory tests and those predicted by each method, for each of the selected voltage waveshapes.

Submitted by Renata Lopes (renatasil82@gmail.com) on 2018-07-13T15:08:29Z No. of bitstreams: 1 thandaramariakathleendasilva.pdf: 8098059 bytes, checksum: 10b3e2aa5918f7826b5f18da04cfae69 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2018-09-03T16:17:06Z (GMT) No. of bitstreams: 1 thandaramariakathleendasilva.pdf: 8098059 bytes, checksum: 10b3e2aa5918f7826b5f18da04cfae69 (MD5) / Made available in DSpace on 2018-09-03T16:17:06Z (GMT). No. of bitstreams: 1 thandaramariakathleendasilva.pdf: 8098059 bytes, checksum: 10b3e2aa5918f7826b5f18da04cfae69 (MD5) Previous issue date: 2018-06-12 / O objetivo deste trabalho é analisar o retorno eleitoral do Programa Minha Casa Minha Vida (MCMV) obtido pela candidata do governo Dilma Rousseff nos pleitos presidenciais. A votação eleitoral no Brasil é influenciada por uma série de características não observadas, tais como clientelismo, coronelismo, preferências individuais, voto de cabresto etc. Se tais características não forem levadas em conta na análise, o retorno eleitoral de programas sociais como MCMV e Bolsa-Família pode não ser adequadamente captado. Para contornar isso, foram estimadas regressões usando dados em painel com controle de efeitos fixos em nível microrregional. Os resultados revelam que há evidência de retorno eleitoral nulo do MCMV. Ao contrário, constatou-se ainda a existência de retorno eleitoral do Bolsa-Família em todas as regressões. / The aim of this study is to analyze the electoral return of the Minha Casa Minha Vida Program (MCMV) obtained by the incumbent candidate Dilma Rousseff in the presidential elections. Brazilian polls are affected by various non-observed characteristics, such as patronage, "coronelismo", individual preferences, "voto de cabresto" etc. If these characteristics are not taken into account in the analysis, the electoral return of social programs like MCMV and Bolsa-Familia (BF) cannot be properly captured. To solve this problem, a fixed effect model was estimated at the micro-regional level. The findings reveal that there is evidence of electoral return null of MCMV. Still, the electoral return of BF program was verified in all estimated regressions.

Rammed earth is a monolithic construction formed by compacting processed soil in progressive layers. Rammed earth is used for the construction of load bearing walls, floors, sub base material in roadways, airport runways, taxiways, aprons, foundations and earthen bunds. Soil, sand, cement and water are the ingredients used for the preparation of cement stabilized rammed earth (CSRE) specimens. The cracking in a rammed earth structure is due to the development of tensile stresses. The tensile stresses are generated due to various causes like unequal settlement of foundation, eccentric loading and / or lateral loading such as wind pressure and earthquake on an earth structure. The cracking in a rammed earth structure causes the failure of its intended function. For example formation of crack may lead to the instability of an embankment slope. And earthen dam can be destroyed gradually by erosion of soil at the crack surface (Harison et al. 1994). Hence, it becomes important to understand the fracture behaviour of cement-stabilized rammed earth structures. Well focused studies in understanding the fracture behaviour of CSRE structures are scanty. The present work attempts to address some issues on the fracture behaviour of CSRE including size effect. Through an experimental programme material properties viz. compressive strength, tensile strength and stress-strain relationships are generated for two chosen densities, 17 and 18.5 kN/m3 of CSRE both in dry and saturated condition. Soil composition, density, cement content and moisture content of the specimen during testing influence the characteristics of CSRE. In the present investigation keeping the cement at 10%, the density is varied choosing a soil-sand mixture having optimum grading limits. The basic raw materials used are soil, sand, cement and water in the ratio of 1 : 1.5 : 0.25 : 0.34 by weight. The strength properties studied alone are inadequate to predict the mechanics of fracture due to the presence of microscopic flaws, cracks, voids and other discontinuities. Therefore, some linear elastic fracture parameters such as mode I fracture toughness (KIc), critical energy release rate (GIc), net section strength (f net) and notch sensitivity are calculated, presuming that CSRE is still a brittle material because it is yet to be confirmed that CSRE is a quasibrittle material. In fact, in the present work, it is shown that CSRE has significant amount of softening. A comprehensive experimental work has been undertaken to test CSRE beam specimens for two densities, three sizes of beam and three notch to depth ratios under three point bending (TPB) in a closed loop servo-controlled machine with crack mouth opening displacement control. Results indicate that the CSRE in dry condition exhibits a greater resistance to fracture than the saturated specimen. The variation of net section strength with the notch depth is not significant. Therefore the CSRE material is notch insensitive, implying that it is less brittle. An experimental program was undertaken to determine the nonlinear fracture parameters of beam specimens both in dry and saturated condition. The influence of moisture content, density, size of the specimen as well as notch to depth ratio of the specimen on RILEM fracture energy (G F ) are presented. The GF values increase with increase in density and size of the specimen, while they decrease with increase in notch to depth ratio. Results clearly show that the total energy absorbed by the beams (W OF ) and RILEM fracture energy (G F ) for all specimens tested in dry state are higher compared to the specimens tested in saturated state, indicating that the dry specimen offers higher resistance to the crack propagation. The RILEM fracture energy GF , determined from TPB tests, is said to be size dependent. The assumption made in the work of fracture is that the total strain energy is utilized for the fracture of the specimen. The fracture energy is proportional to the size of the fracture process zone (FPZ), which also implies that size of FPZ increases with increase in the un-cracked ligament (d - a) of beam. This also means that FPZ is proportional to the depth d for a given notch to depth ratio, because for a given notch/depth, (d - a) which is also is proportional to d because is a constant. This corroborates the fact that fracture energy increases with size. Interestingly, the same conclusion has been drawn by Karihaloo et al. (2006). They have plotted a curve relating fracture process zone length and overall depth the beam. In the present study a new method namely Fracture energy release rate method proposed by Muralidhara et al. (2013) is used. In the new method the plot of GF /(d - a) versus (d - a) is obtained from a set of experimental results. The plot is found to follow power law and showed almost constant value of GF /(d - a) at larger ligament lengths. This means the fracture energy reaches a constant value at large ligament lengths reaffirming that the fracture energy from very large specimen is size-independent. This Fracture energy release rate method is used to determine size-independent fracture energy GRf , based on the relationship between RILEM fracture energy and the un-cracked ligament length. The experimental results from the present work agree well with the proposed new method. Similarly, the method is extended to determine nominal shear strength τv for large size beam. Results show that for both densities GRf decrease in saturated condition, while in dry condition as the density is increased from 17 to 18.5 kN/m3 the GRf decrease by 7.58%, indicating that the brittleness increases with higher density. The τv for large size beam increases with density both in dry and saturated condition. The size effect method for evaluating material fracture properties proposed by Bazant (1984) is applied to cement stabilised rammed earth. By measuring the peak loads of 2D geometrically similar notched beam specimens of different sizes, nonlinear fracture parameters such as fracture energy (Gf ), fracture toughness (KIc), effective length of the fracture process zone (Cf ), brittleness number (β), characteristic length (l 0) and the critical crack tip opening displacement (CT ODc) are determined for both dry and saturated conditions. The crack growth resistance curves (R-curve) are also developed for dry and saturated specimens. In the size effect method, for both densities 18.5 and 17 kN/m3 the values of nonlinear fracture properties, namely G f , Cf , KIc, CT ODc and l 0 are lower for the saturated specimen compared to those of the dry specimen. In dry condition as the density is increased from 17 to 18.5 kN/m3 the Gf decreases to 13.54%, indicating that the brittleness increase with higher density. The areas under the load-displacement and load-CMOD curves are a measure of the fracture energy and these areas are low for saturated specimens. The crack growth resistance curves (R-curve) plotted using the size-effect law from peak loads are the measure of resistance against crack growth R. The value of R is high for dry specimen compared to that of the saturated specimens. During aggregate pullout or the opening of crack, the interlock or friction between the crack surfaces may cause the energy dissipation through friction and bridging across the crack. Therefore the wet friction in case of saturated specimen must be smaller resulting in more brittleness compared to the larger dry friction for dry specimen. In the present investigation the Digital Image Correlation (DIC) technique is used to study the FPZ properties in cement stabilised rammed earth. The MATLAB package written by Eberl et al. (2006) is suitably modified and used for image correlation to suit our requirements. CMOD measured using DIC technique is validated by comparison with the CMOD measured using clip gauge. The FPZ properties such as the development of FPZ and crack opening displacements at different loading points as well as the influence of notch/depth ratio on FPZ length (lFPZ ) are evaluated for both dry and saturated conditions. At peak load the lFPZ are about 0.315 and 0.137 times the un-cracked ligament length respectively for specimens tested under dry and saturated conditions. In dry and saturated states the FPZ length decreases as the ratio increases. Lower values of lFPZ in saturated specimen indicates that it is relatively more brittle compared to dry specimen.

La découverte de nouveaux médicaments est un processus complexe qui implique de gros investissements et de longues phases de développement. Le coût des échecs est phénoménal et le taux de succès diminue depuis deux décennies, suggérant que le processus d'innovation-développement tel qu'il est pratiqué aujourd'hui n'est plus adapté aux défis actuels. Dans ce travail de thèse, nous proposons une nouvelle approche, dite descendante « Top-down » qui consiste à inverser le procédé traditionnel. Elle reposera essentiellement sur la conception et l'optimisation de modèles de prédiction, afin de mieux orienter les phases successives du développement et de prédire les résultats de chaque étape avec le maximum de précision. Cette démarche s'appuie sur la conception du modèle d'innovation thérapeutique et de toutes ses composantes, notamment le modèle physiopathologique (modèle simulant l'organisme et l'évolution naturelle de la maladie) le modèle thérapeutique (modèle simulant l'effet du médicament sur la maladie), et le modèle d'effet (modèle qui peut être prédit à partir des modèles précédents et qui permet de simuler l'impact du médicament sur la population traitée à partir de son effet sur chaque individu de la population. Le processus complet de notre stratégie comprend dix étapes : élaboration du modèle physiopathologique, calibration, validation, identification des cibles potentielles, activation du logiciel de modifications séquentielles des cibles, construction de la population virtuelle (réaliste ou non), obtention du modèle d'effet nécessaire pour calculer le nombre d'événements évités (NEE), calcul du NEE par cible, comparaison et choix de la “meilleure” cible / New drug discovery and development is a complex process which requires massive investments over protracted horizons. The cost of failed programs is significant in absolute dollar terms. Decreasing R&D productivity over the past 2 decades suggests that the traditional innovation model needs a radical rethink. The central purpose of this thesis work is to lay the theoretical foundations for a new "topdown" approach to new target identification construed as an alternative to the current bottomup approach based on high throughput screening. In essence, it will consist in the design and optimization of in silico models in order to guide the development of a new drug candidate by predicting the outcomes of each successive phase of the development process. This in silico framework brings together a physiopathological model (to simulate the natural evolution of the disease) with a therapeutic model (to simulate the effects of a drug candidate on disease evolution) and the effect model (to predict the impact of the drug candidate on a population of patients)