The aim of the thesis is to evaluate an investment project in the Russian Federation from the foreign investor's perspective. The project feasibility study was developed as the first step in a decision making process of execution an international investment project, which includes ambient conditions and market analysis. The plan of cash flow was compiled in compliance of Russian tax system, and was considered as a base for project evaluation by using appropriate methods. For further assessment of possible risks of the project was used sensitivity analysis.

À mesure que les performances intrinsèques des bâtiments s’améliorent, les usages énergétiques non réglementés, que nous associons à une notion d’intensité énergétique des usages, prennent de plus en plus d’importance dans le bilan des consommations des bâtiments. De plus, les bâtiments performants font apparaître des problématiques au niveau de l’exploitation des installations. Ces constats nous permettent d’affirmer qu’il est aujourd’hui important de proposer un cadre pour le suivi et l’optimisation de la sobriété énergétique des usages et l’exploitation performante pour la maîtrise des consommations énergétiques réelles des bâtiments. Cette thèse propose tout d’abord de développer des modèles polynomiaux de prédiction de la consommation énergétique tous usages en fonction des facteurs caractérisant l’intensité d’usage, la qualité d’usage et la qualité d’exploitation. Pour cela, nous utilisons le logiciel EnergyPlus afin de réaliser des simulations énergétiques dynamiques (SED) sur des valeurs de paramètres définis par la méthode des plans d’expérience D-optimaux. Le modèle polynomial créé permet alors d’effectuer, avec un faible temps de calcul, une propagation des incertitudes sur les consommations d’énergie calculées. Pour ce faire, nous utilisons les données mesurées en exploitation dans le cadre de la mesure et de la vérification de la performance énergétique, associées à une incertitude concernant leur valeur. Nous pouvons alors déterminer l’incertitude globale sur les consommations énergétiques et identifier les pistes pour la réduire, permettant ainsi un meilleur suivi et encadrement de la consommation énergétique réelle. / Since building envelope and MEP systems characteristics regularly improve, the weight of non-regulatory energy end-uses increases. These energy end-uses are typically associated with tenants or owners’ activities. In addition, high performance buildings show new issues related to HVAC systems operations. Therefore, it is important to evaluate and improve non-regulatory energy end-uses energy as well as HVAC systems operations efficiencies. We have developed polynomial energy models that can predict energy consumption as a function of building’s activities characteristics and HVAC systems operations factors. We used EnergyPlus software in order to build reliable energy models along with the D-optimum design of experiments method (DOE). Then, we used measurement and verification (M&V) data, associated with probability functions, to determine the associated uncertainty of the calculated energy consumption. Finally, we combine the latter with the polynomial modeling error to calculate the energy consumption global uncertainty, with the goal to identify strategies to reduce it.

Nesta tese apresentam-se desenvolvimentos metodológicos para analisar dados com omissão e também estudos delineados para compreender os resultados de tais análises. Escrutinam-se análises de sensibilidade bayesiana e clássica para dados com respostas categorizadas sujeitas a omissão. Mostra-se que as componentes subjetivas de cada abordagem podem influenciar os resultados de maneira não-trivial, independentemente do tamanho da amostra, e que, portanto, as conclusões devem ser cuidadosamente avaliadas. Especificamente, demonstra-se que distribuições \\apriori\\ comumente consideradas como não-informativas ou levemente informativas podem, na verdade, ser bastante informativas para parâmetros inidentificáveis, e que a escolha do modelo sobreparametrizado também tem um papel importante. Quando há omissão em variáveis explicativas, também é necessário propor um modelo marginal para as covariáveis mesmo se houver interesse apenas no modelo condicional. A especificação incorreta do modelo para as covariáveis ou do modelo para o mecanismo de omissão leva a inferências enviesadas para o modelo de interesse. Trabalhos anteriormente publicados têm-se dividido em duas vertentes: ou utilizam distribuições semiparamétricas/não-paramétricas, flexíveis para as covariáveis, e identificam o modelo com a suposição de um mecanismo de omissão não-informativa, ou empregam distribuições paramétricas para as covariáveis e permitem um mecanismo mais geral, de omissão informativa. Neste trabalho analisam-se respostas binárias, combinando um mecanismo de omissão informativa com um modelo não-paramétrico para as covariáveis contínuas, por meio de uma mistura induzida pela distribuição \\apriori\\ de processo de Dirichlet. No caso em que o interesse recai apenas em momentos da distribuição das respostas, propõe-se uma nova análise de sensibilidade sob o enfoque clássico para respostas incompletas que evita suposições distribucionais e utiliza parâmetros de sensibilidade de fácil interpretação. O procedimento tem, em particular, grande apelo na análise de dados contínuos, campo que tradicionalmente emprega suposições de normalidade e/ou utiliza parâmetros de sensibilidade de difícil interpretação. Todas as análises são ilustradas com conjuntos de dados reais. / We present methodological developments to conduct analyses with missing data and also studies designed to understand the results of such analyses. We examine Bayesian and classical sensitivity analyses for data with missing categorical responses and show that the subjective components of each approach can influence results in non-trivial ways, irrespectively of the sample size, concluding that they need to be carefully evaluated. Specifically, we show that prior distributions commonly regarded as slightly informative or non-informative may actually be too informative for non-identifiable parameters, and that the choice of over-parameterized models may drastically impact the results. When there is missingness in explanatory variables, we also need to consider a marginal model for the covariates even if the interest lies only on the conditional model. An incorrect specification of either the model for the covariates or of the model for the missingness mechanism leads to biased inferences for the parameters of interest. Previously published works are commonly divided into two streams: either they use semi-/non-parametric flexible distributions for the covariates and identify the model via a non-informative missingness mechanism, or they employ parametric distributions for the covariates and allow a more general informative missingness mechanism. We consider the analysis of binary responses, combining an informative missingness model with a non-parametric model for the continuous covariates via a Dirichlet process mixture. When the interest lies only in moments of the response distribution, we consider a new classical sensitivity analysis for incomplete responses that avoids distributional assumptions and employs easily interpreted sensitivity parameters. The procedure is particularly useful for analyses of missing continuous data, an area where normality is traditionally assumed and/or relies on hard-to-interpret sensitivity parameters. We illustrate all analyses with real data sets.

Des analyses de sensibilité et des estimations de paramètres sont étudiées sur deux études de cas de transfert de masse en milieu poreux. La première partie est consacrée à la sensibilité des écoulements souterrains dans une modélisation des échanges drain-aquifère pour mettre en évidence les différences entre les deux méthodes de discrétisation mises en œuvre. La seconde partie est dédiée à la modélisation de l’écoulement en milieu poreux variablement saturé dans une zone humide artificielle, au calage des paramètres du modèle de van Genuchten-Mualem et à l’évaluation de son efficacité à reproduire des données piézométriques collectées sur le site de l’Ostwaldergraben. La variabilité temporelle des paramètres hydrodynamiques, incluant l’effet d’hystérésis, montre que ceux de la couche active du filtre changent au cours du temps. Ces deux études sont conduites à l’aide de la différenciation automatique. / Sensitivity analyses and parameter estimation are applied to mass transfer in porous media for two remediation facilities. The first part is devoted to the sensitivity analysis of groundwater flows in a modeling of drain-aquifer exchanges to highlight the differences between the two implemented methods of discretization. The second part is dedicated to the modeling of the flow in a variably saturated porous medium in a stormwater constructed wetland, to the calibration of van Genuchten-Mualem parameters and to the evaluation of its efficiency in the reproduction of piezometric data collected on the Ostwaldergraben site. The temporal variability of the hydrodynamic parameters, including the hysteresis effect, shows that the characteristics of the filter layer alters along time. Both studies are carried using automatic differentiation.

Les épilepsies acquises se déclarent après un processus graduel appelé épileptogenèse. Bien que cliniquement silencieux, ce processus implique des modifications fonctionnelles observables notamment par électroencéphalographie. Cette thèse vise i) à identifier des marqueurs électrophysiologiques apparaissant au cours de l’épileptogenèse, et ii) à comprendre les modifications physiopathologiques sous-jacentes responsables de ces marqueurs et de leur évolution temporelle. Dans un premier temps, nous avons, dans un modèle d’épilepsie partielle chez la souris, monitoré des signaux électrophysiologiques intracérébraux pendant la mise en place de la maladie. Nous avons observé dans ces signaux expérimentaux, l’émergence d’événements transitoires pathologiques appelés pointes épileptiques. Nous avons développé des méthodes de traitement du signal pour détecter et caractériser automatiquement ces événements. Ainsi, nous avons pu mettre en évidence certains changements dans la forme des pointes épileptiques au cours de l’épileptogenèse ; en particulier l’apparition et l’augmentation d’une onde qui suit la pointe épileptique. Une hypothèse défendue dans ces travaux est que ces changements morphologiques peuvent constituer des marqueurs de l’épileptogenèse dans ce modèle animal. Dans un second temps, afin d’interpréter ces modifications électrophysiologiques en termes de processus neurophysiologiques sous-jacents, nous avons implémenté un modèle biomathématique, physiologiquement argumenté, capable de simuler des pointes épileptiques. Formellement, ce modèle est un système dynamique non linéaire qui reproduit les interactions synaptiques (excitatrices et inhibitrices) dans une population de neurones. Une analyse de sensibilité de ce modèle a permis de mettre en évidence le rôle critique de certains paramètres de connectivité dans la morphologie des pointes. Nos résultats montrent en effet, qu’une diminution de l’inhibition GABAergique entraîne un accroissement de l’onde dans les pointes épileptiques. À partir du modèle théorique, nous avons pu ainsi émettre des hypothèses sur les modifications opérant au cours du processus d’épileptogenèse. Ces hypothèses ont pu être en partie vérifiées expérimentalement en bloquant artificiellement l’inhibition GABAergique, dans le modèle in vivo chez la souris, et dans un modèle in vitro chez le rat. En conclusion, ce travail de thèse fournit, dans un modèle animal, un biomarqueur électrophysiologique de l’épileptogenèse et tente d’expliquer, grâce à une modélisation biomathématique, les processus neurophysiologiques sous-jacents qu’il reflète. / Acquired epilepsies occur after a process called epileptogenesis. Although clinically silent, this process involves some functional modifications which can be observed by electroencephalography. The objectives of this thesis are i) to identify electrophysiological markers occurring during epileptogenesis, and ii) to understand which underlying pathophysiological modifications are responsible for these markers and their evolution. Firstly, using an in vivo experimental mouse model of partial epilepsy, we have monitored intracranial electrophysiological signals during epileptogenesis. We observed the emergence of pathological transient events called epileptic spikes. We have developed signal processing methods in order to automatically detect and characterize these events. Hence, we observed and quantified morphological changes of epileptic spikes during epileptogenesis. In particular, we noticed the emergence and the increase of a wave which directly follows the spike component. In this work, we defend the hypothesis that these morphological modifications can constitute markers of the epileptogenesis process in this animal model of epilepsy. Secondly, in order to interpret these electrophysiological modifications in terms of underlying pathophysiological processes, we have implemented a computational model able to simulate epileptic spikes. This neural mass model is a neurophysiologically-plausible mesoscopic representation of synaptic interactions (excitation and inhibition) in the hippocampus. Based on a sensitivity analysis of model parameters, we were able to determine some connectivity parameters that play a key role in the morphology of simulated epileptic spikes. In particular, our results show that a diminution of GABAergic inhibition leads to an increase of the aforementioned wave. Thus, using this theoretical model, we defined some hypotheses about pathophysiological modifications occurring during the epileptogenesis process. One of these hypotheses has been confirmed in blocking GABAa receptors in the in vivo mouse model, as well as in an in vitro model (rat, organotypic slices). In summary, based on the shape features of epileptic spikes, we devised an electrophysiological biomarker of epileptogenesis observed in a mouse model but useful in Human studies as well. Moreover, a computational modeling approach has permitted to suggest which pathophysiological processes might underlie this biomarker.

Nesta tese apresentam-se desenvolvimentos metodológicos para analisar dados com omissão e também estudos delineados para compreender os resultados de tais análises. Escrutinam-se análises de sensibilidade bayesiana e clássica para dados com respostas categorizadas sujeitas a omissão. Mostra-se que as componentes subjetivas de cada abordagem podem influenciar os resultados de maneira não-trivial, independentemente do tamanho da amostra, e que, portanto, as conclusões devem ser cuidadosamente avaliadas. Especificamente, demonstra-se que distribuições \\apriori\\ comumente consideradas como não-informativas ou levemente informativas podem, na verdade, ser bastante informativas para parâmetros inidentificáveis, e que a escolha do modelo sobreparametrizado também tem um papel importante. Quando há omissão em variáveis explicativas, também é necessário propor um modelo marginal para as covariáveis mesmo se houver interesse apenas no modelo condicional. A especificação incorreta do modelo para as covariáveis ou do modelo para o mecanismo de omissão leva a inferências enviesadas para o modelo de interesse. Trabalhos anteriormente publicados têm-se dividido em duas vertentes: ou utilizam distribuições semiparamétricas/não-paramétricas, flexíveis para as covariáveis, e identificam o modelo com a suposição de um mecanismo de omissão não-informativa, ou empregam distribuições paramétricas para as covariáveis e permitem um mecanismo mais geral, de omissão informativa. Neste trabalho analisam-se respostas binárias, combinando um mecanismo de omissão informativa com um modelo não-paramétrico para as covariáveis contínuas, por meio de uma mistura induzida pela distribuição \\apriori\\ de processo de Dirichlet. No caso em que o interesse recai apenas em momentos da distribuição das respostas, propõe-se uma nova análise de sensibilidade sob o enfoque clássico para respostas incompletas que evita suposições distribucionais e utiliza parâmetros de sensibilidade de fácil interpretação. O procedimento tem, em particular, grande apelo na análise de dados contínuos, campo que tradicionalmente emprega suposições de normalidade e/ou utiliza parâmetros de sensibilidade de difícil interpretação. Todas as análises são ilustradas com conjuntos de dados reais. / We present methodological developments to conduct analyses with missing data and also studies designed to understand the results of such analyses. We examine Bayesian and classical sensitivity analyses for data with missing categorical responses and show that the subjective components of each approach can influence results in non-trivial ways, irrespectively of the sample size, concluding that they need to be carefully evaluated. Specifically, we show that prior distributions commonly regarded as slightly informative or non-informative may actually be too informative for non-identifiable parameters, and that the choice of over-parameterized models may drastically impact the results. When there is missingness in explanatory variables, we also need to consider a marginal model for the covariates even if the interest lies only on the conditional model. An incorrect specification of either the model for the covariates or of the model for the missingness mechanism leads to biased inferences for the parameters of interest. Previously published works are commonly divided into two streams: either they use semi-/non-parametric flexible distributions for the covariates and identify the model via a non-informative missingness mechanism, or they employ parametric distributions for the covariates and allow a more general informative missingness mechanism. We consider the analysis of binary responses, combining an informative missingness model with a non-parametric model for the continuous covariates via a Dirichlet process mixture. When the interest lies only in moments of the response distribution, we consider a new classical sensitivity analysis for incomplete responses that avoids distributional assumptions and employs easily interpreted sensitivity parameters. The procedure is particularly useful for analyses of missing continuous data, an area where normality is traditionally assumed and/or relies on hard-to-interpret sensitivity parameters. We illustrate all analyses with real data sets.

本研究的主要目的是檢視倒傳導神經網路是否具有人類在分類學習上所呈現出來的學習效應 — 競爭學習、遮蔽效應與不相關線索的影響。在實驗中，我們採用兩種倒傳導神經網路，來測試激發函數是否會影響倒傳導神經網路的學習。此兩種倒傳導神經網路分別採用sigmoid激發函數與hyperbolic-tangent激發函數。實驗結果顯示，以sigmoid為激發函數與以hyperbolic-tangent為激發函數的倒傳導神經網路都具有這三個學習效應。還有，以sigmoid為激發函數的倒傳導神經網路所呈現出來的學習效應比以hyperbolic-tangent為激發函數的倒傳導神經網路來得顯著。本研究的次要目的在於瞭解有效性（使用性）與敏感度分析的數值是否有對應關係。實驗結果顯示，線索A與線索B的敏感度分析數值差異可以反映出線索A與線索B的有效性差異。然而，敏感度分析數值卻無法準確地顯示線索的有效性數值。 / The main objective of this research is to examine whether back propagation neural networks (BP) have the learning effects found in human category learning — competitive learning, overshadowing and the deleterious of an irrelevant cue. Two kinds of BP, BP with sigmoid activation function and BP with hyperbolic-tangent activation function, are investigated to see if the activation function will make BP behave differently. According to the results of our experiments, these three learning effects are demonstrated both in BP with sigmoid and BP with hyperbolic-tangent, but they seems more significant in BP with sigmoid than in BP with hyperbolic-tangent. The second objective of our research is to see if there is a correspondence between the validity (the utilization) and the value of sensitivity analysis, R. From the results of our experiments, we observe that the difference between values of sensitivity analysis with respect to Cue A and Cue B reflects the difference of the validities between Cue A and Cue B. However, the value of sensitivity analysis does not show exactly what validity a cue is.

本篇論文旨在分析神經網路學習績效，並提出一套學習演算法，結合倒傳遞網路(BP)與理解神經網路(RN)，命名為RNBP，這套學習演算法將與傳統的BP做比較，以兩個不同的財務金融領域的應用，一個是選擇權上Black-Scholes訂價模式的模擬，一個是金融交換上利率的預測。主要績效的評估準則是以學習的效率與模擬、預測的準確度為依據。 此外，本論文的另一個重點是提出一套對於神經網路系統進一步分析的方法與工具，敏感度分析(Sensitivity Analysis)與滯留區(Dead Region)分析，藉以瞭解神經網路系統是否具有效地良好學習或被一般化的能力，從神經網路的角度來說，這也是BP與RNBP的另一個績效比較標準。本研究的結果顯示RNBP在預測準確度上較BP為優良，但是在學習效率與預測能力的穩定性上並沒有呈現一致性的結論；此外，敏感度分析與滯留區分析的結果也幫助神經網路在應用領域上有更深入的瞭解。 在過去，神經網路的應用者往往忽略了進一步瞭解神經網路的重要性與可行性，本論文的貢獻在於藉由分析神經網路所學習的知識，幫助應用者進一步瞭解神經網路表達的訊息在應用領域上所隱含的實質意義。 / The study attempts to analyze the learning performance of neural networks in applications, and propose a new learning procedure for the layered feedforward neural network systems, named KNBP, which binds RN and BP learning algorithms. Two artificial neural networks, BP and KNBP, here are both applied to two financial fields, the simulation of Black-Scholes pricing model for the call options and the midrates forecasting in financial swaps. The explicit performance comparison between the two artificial neural network systems is mainly based on two criteria, which are learning efficiency and forecasting effectiveness. Then we propound a mathematical methodology of sensitivity analysis and the dead regions to deeply explore inside the network structures to see whether the models of ANNS are actually well trained or valid, and thus setup an alternative comparable criterion. The results from this study show that RNBP performs better than BP in forecasting effectiveness, but RNBP obtains neither a consistent learning efficiency in cases nor a stable forecasting ability. Furthermore, the sensitivity analysis and the dead region analysis provide a deeper view of the ANNs in the applied fields. In the past, most studies applying neural networks ignored the importance that it is feasible and advantageous to obtain more useful information via analyzing neural networks. The purpose of the research is to help further understanding to the information discovery resulted from neural networks in practical applications.

La modélisation du vivant, en particulier la modélisation de l'activité cardiaque, est devenue un défi scientifique majeur. Le but de cette thématique est de mieux comprendre les phénomènes physiologiques et donc d'apporter des solutions à des problèmes cliniques. Nous nous intéressons dans cette thèse à la modélisation et à l'étude numérique de l'activité électrique du cœur, en particulier l'étude des électrocardiogrammes (ECGs). L'onde électrique dans le cœur est gouvernée par un système d'équations de réaction-diffusion appelé modèle bidomaine ce système est couplé à une EDO représentant l'activité cellulaire. Afin simuler des ECGs, nous tenons en compte la propagation de l'onde électrique dans le thorax qui est décrite par une équation de diffusion. Nous commençons par une démonstrer l'existence d'une solution faible du système couplé cœur-thorax pour une classe de modèles ioniques phénoménologiques. Nous prouvons ensuite l'unicité de cette solution sous certaines conditions. Le plus grand apport de cette thèse est l'étude et la simulation numérique du couplage électrique cœur-thorax. Les résultats de simulations sont représentés à l'aide des ECGs. Dans une première partie, nous produisons des simulations pour un cas normal et pour des cas pathologiques (blocs de branche gauche et droit et des arhythmies). Nous étudions également l'impact de certaines hypothèses de modélisation sur les ECGs (couplage faible, utilisation du modèle monodomaine, isotropie, homogénéité cellulaire, comportement résistance-condensateur du péricarde,. . . ). Nous étudions à la fin de cette partie la sensibilité des ECGs par apport aux paramètres du modèle. En deuxième partie, nous effectuons l'analyse numérique de schémas du premier ordre en temps découplant les calculs du potentiel d'action et du potentiel extérieur. Puis, nous combinons ces schémas en temps avec un traîtement explicite du type Robin-Robin des conditions de couplage entre le cœur et le thorax. Nous proposons une analyse de stabilité de ces schémas et nous illustrons les résultats avec des simulations numériques d'ECGs. La dernière partie est consacrée à trois applications. Nous commençons par l'estimation de certains paramètres du modèle (conductivité du thorax et paramètres ioniques). Dans la deuxième application, qui est d'originie industrielle, nous utilisons des méthodes d'apprentissage statistique pour reconstruire des ECGs à partir de mesures ('électrogrammes). Enfin, nous présentons des simulations électro-mécaniques du coeur sur une géométrie réelle dans diverses situations physiologiques et pathologiques. Les indicateurs cliniques, électriques et mécaniques, calculés à partir de ces simulations sont très similaires à ceux observés en réalité.

<p>Accelerator-driven systems for transmutation of nuclear waste have been suggested as a means for dealing with spent fuel components that pose potential radiological hazard for long periods of time. While not entirely removing the need for underground waste repositories, this nuclear waste incineration technology provides a viable method for reducing both waste volumes and storage times. Potentially, the time spans could be diminished from hundreds of thousand years to merely 1.000 years or even less. A central aspect for accelerator-driven systems design is the prediction of safety parameters and fuel economy. The simulations performed rely heavily on nuclear data and especially on the precision of the neutron cross section representations of essential nuclides over a wide energy range, from the thermal to the fast energy regime. In combination with a more demanding neutron flux distribution as compared with ordinary light-water reactors, the expanded nuclear data energy regime makes exploration of the cross section sensitivity for simulations of accelerator-driven systems a necessity. This fact was observed throughout the work and a significant portion of the study is devoted to investigations of nuclear data related effects. The computer code package EA-MC, based on 3-D Monte Carlo techniques, is the main computational tool employed for the analyses presented. Directly related to the development of the code is the extensive IAEA ADS Benchmark 3.2, and an account of the results of the benchmark exercises as implemented with EA-MC is given. CERN's Energy Amplifier prototype is studied from the perspectives of neutron source types, nuclear data sensitivity and transmutation. The commissioning of the n_TOF experiment, which is a neutron cross section measurement project at CERN, is also described.</p>