La toxicité et l’efficacité de plus de 30 agents anticancéreux présentent de très fortes variations en fonction du temps de dosage. Par conséquent, les biologistes qui étudient le rythme circadien ont besoin d’une méthode très précise pour estimer le vecteur de composantes périodiques (CP) de signaux chronobiologiques. En outre, dans les développements récents, non seulement la période dominante ou le vecteur de CP présentent un intérêt crucial, mais aussi leurs stabilités ou variabilités. Dans les expériences effectuées en traitement du cancer, les signaux enregistrés correspondant à différentes phases de traitement sont courts, de sept jours pour le segment de synchronisation jusqu’à deux ou trois jours pour le segment après traitement. Lorsqu’on étudie la stabilité de la période dominante nous devons considérer des signaux très court par rapport à la connaissance a priori de la période dominante, placée dans le domaine circadien. Les approches classiques fondées sur la transformée de Fourier (TF) sont inefficaces (i.e. manque de précision) compte tenu de la particularité des données (i.e. la courte longueur). Dans cette thèse, nous proposons une nouvelle méthode pour l’estimation du vecteur de CP des signaux biomédicaux, en utilisant les informations biologiques a priori et en considérant un modèle qui représente le bruit. Les signaux enregistrés dans le cadre d’expériences développées pour le traitement du cancer ont un nombre limité de périodes. Cette information a priori peut être traduite comme la parcimonie du vecteur de CP. La méthode proposée considère l’estimation de vecteur de CP comme un problème inverse enutilisant l’inférence bayésienne générale afin de déduire toutes les inconnues de notre modèle, à savoir le vecteur de CP mais aussi les hyperparamètres (i.e. les variances associées). / The toxicity and efficacy of more than 30 anticancer agents presents very high variations, depending on the dosing time. Therefore the biologists studying the circadian rhythm require a very precise method for estimating the Periodic Components (PC) vector of chronobiological signals. Moreover, in recent developments not only the dominant period or the PC vector present a crucial interest, but also their stability or variability. In cancer treatment experiments the recorded signals corresponding to different phases of treatment are short, from seven days for the synchronization segment to two or three days for the after treatment segment. When studying the stability of the dominant period we have to consider very short length signals relative to the prior knowledge of the dominant period, placed in the circadian domain. The classical approaches, based on Fourier Transform (FT) methods are inefficient (i.e. lack of precision) considering the particularities of the data (i.e. the short length). In this thesis we propose a new method for the estimation of the PC vector of biomedical signals, using the biological prior informations and considering a model that accounts for the noise. The experiments developed in the cancer treatment context are recording signals expressing a limited number of periods. This is a prior information that can be translated as the sparsity of the PC vector. The proposed method considers the PC vector estimation as an Inverse Problem (IP) using the general Bayesian inference in order to infer all the unknowns of our model, i.e. the PC vector but also the hyperparameters.

The main objective of my diploma thesis is to analyze and evaluate the Political Movement ANO positioning within the political parties system of the Czech Republic by using a methodological framework approach. The thesis provides an analysis of electorate shifting and selected political parties manifestos as well as their comparison with the Political Movement ANO. Timewise, my focus is on the period prior to the election to the Chamber of Deputies of the Parliament of the Czech Republic in 2017. As for analytical purposes, the Hierarchical Bayesian Modeling has been used. This statistical model helps to get the respective values and to show the electoral vote changes between the last two parliament elections (to Chamber of Deputies). The author uses quantitative and qualitative research for comparison and analysis of programmatical convergency as defined in the election manifestos of various political parties. For manifestos quantification a coding scheme by a Comparative manifesto project group has been applied. The reason for using the above mentioned scheme is that it provides a structured methodology to quantify the domains that the political parties do focus the most in their manifestos. The aim of the analytical part of the thesis is to define how and especially from where the Movement ANO...

No description available.

In the thesis "On Boundaries of Statistical Models" problems related to a description of probability distributions with zeros, lying in the boundary of a statistical model, are treated. The distributions considered are joint distributions of finite collections of finite discrete random variables. Owing to this restriction, statistical models are subsets of finite dimensional real vector spaces. The support set problem for exponential families, the main class of models considered in the thesis, is to characterize the possible supports of distributions in the boundaries of these statistical models. It is shown that this problem is equivalent to a characterization of the face lattice of a convex polytope, called the convex support. The main tool for treating questions related to the boundary are implicit representations. Exponential families are shown to be sets of solutions of binomial equations, connected to an underlying combinatorial structure, called oriented matroid. Under an additional assumption these equations are polynomial and one is placed in the setting of commutative algebra and algebraic geometry. In this case one recovers results from algebraic statistics. The combinatorial theory of exponential families using oriented matroids makes the established connection between an exponential family and its convex support completely natural: Both are derived from the same oriented matroid. The second part of the thesis deals with hierarchical models, which are a special class of exponential families constructed from simplicial complexes. The main technical tool for their treatment in this thesis are so called elementary circuits. After their introduction, they are used to derive properties of the implicit representations of hierarchical models. Each elementary circuit gives an equation holding on the hierarchical model, and these equations are shown to be the "simplest", in the sense that the smallest degree among the equations corresponding to elementary circuits gives a lower bound on the degree of all equations characterizing the model. Translating this result back to polyhedral geometry yields a neighborliness property of marginal polytopes, the convex supports of hierarchical models. Elementary circuits of small support are related to independence statements holding between the random variables whose joint distributions the hierarchical model describes. Models for which the complete set of circuits consists of elementary circuits are shown to be described by totally unimodular matrices. The thesis also contains an analysis of the case of binary random variables. In this special situation, marginal polytopes can be represented as the convex hulls of linear codes. Among the results here is a classification of full-dimensional linear code polytopes in terms of their subgroups. If represented by polynomial equations, exponential families are the varieties of binomial prime ideals. The third part of the thesis describes tools to treat models defined by not necessarily prime binomial ideals. It follows from Eisenbud and Sturmfels'' results on binomial ideals that these models are unions of exponential families, and apart from solving the support set problem for each of these, one is faced with finding the decomposition. The thesis discusses algorithms for specialized treatment of binomial ideals, exploiting their combinatorial nature. The provided software package Binomials.m2 is shown to be able to compute very large primary decompositions, yielding a counterexample to a recent conjecture in algebraic statistics.

<p>Postpartum depression has become a major public health concern for women within a specific time period after delivery. Depression is possibly associated with some risk factors such as socioeconomic status, social support, maternal mental and physical health, and history of anxiety. TOMIS III, funded by the Canadian Institutes of Health Research, is a prospective cohort to study the associations between delivery method and health and health resource utilization.</p> <p>Clinically, we investigated the associations between mode of delivery and outcome of postnatal depression, maternal and infant health, and we implied the risk predictors for outcomes by statistical methodology of marginal model with generalized estimating equations (GEE). Statistically, a variety of regression models, namely, generalized linear mixed effect model (GLMM), hierarchical generalized linear model (HGLM) and Bayesian hierarchical model were applied for this analysis and results were compared with GEEs. Some imputation strategies, namely, mean imputation, last observation carrying forward (LOCF), hot-deck imputation and multiple imputation were employed for handling missing values in this study.</p> <p>Analysis results demonstrated that there was no statistically significant association between mode of delivery and postpartum depression [OR 0.99, 95% CI (0.73, 1.34)]. However, the development of postpartum depression was found to be associated with low income, low mental and physical health functioning, lack of social support, the low number of unmet learning needs in hospital, and English or French spoken at home. Results were consistent for all regression models but GEE provided the best fit and an excellent discriminative ability. GEE models were constructed on different datasets imputed by mean, LOCF, hot-deck and multiple imputation, and LOCF was recommended to handle the missing data in this longitudinal study.</p> <p>Analyses on the outcome of maternal health and infant health stated that method of delivery had a statistically significant influence on maternal health but no significant impact on infant health. Risks of maternal health problems were associated with cesarean delivery, good/fair/poor infant health, low maternal mental and physical health functioning, lack of care for maternal mental health, and good/fair/poor health before pregnancy. Risks of infant health problems were associated with good/fair/poor maternal health before pregnancy and after discharge, inadequate care or help for infant health, fair/poor community services after discharge, low maternal mental health functioning, non-English or non-French spoken at home, and mothers born outside of Canada.</p> / Master of Science (MSc)

A variety of complex data has broadened in many research fields such as epidemiology, genomics, and analytical chemistry with the development of science, technologies, and design scheme over the past few decades. For example, in epidemiology, the matched case-crossover study design is used to investigate the association between the clustered binary outcomes of disease and a measurement error in covariate within a certain period by stratifying subjects' conditions. In genomics, high-correlated and high-dimensional(HCHD) data are required to identify important genes and their interaction effect over diseases. In analytical chemistry, multiple time series data are generated to recognize the complex patterns among multiple classes. Due to the great diversity, we encounter three problems in analyzing those complex data in this dissertation. We have then provided several contributions to semiparametric and nonparametric methods for dealing with the following problems: the first is to propose a method for testing the significance of a functional association under the matched study; the second is to develop a method to simultaneously identify important variables and build a network in HDHC data; the third is to propose a multi-class dynamic model for recognizing a pattern in the time-trend analysis. For the first topic, we propose a semiparametric omnibus test for testing the significance of a functional association between the clustered binary outcomes and covariates with measurement error by taking into account the effect modification of matching covariates. We develop a flexible omnibus test for testing purposes without a specific alternative form of a hypothesis. The advantages of our omnibus test are demonstrated through simulation studies and 1-4 bidirectional matched data analyses from an epidemiology study. For the second topic, we propose a joint semiparametric kernel machine network approach to provide a connection between variable selection and network estimation. Our approach is a unified and integrated method that can simultaneously identify important variables and build a network among them. We develop our approach under a semiparametric kernel machine regression framework, which can allow for the possibility that each variable might be nonlinear and is likely to interact with each other in a complicated way. We demonstrate our approach using simulation studies and real application on genetic pathway analysis. Lastly, for the third project, we propose a Bayesian focal-area detection method for a multi-class dynamic model under a Bayesian hierarchical framework. Two-step Bayesian sequential procedures are developed to estimate patterns and detect focal intervals, which can be used for gas chromatography. We demonstrate the performance of our proposed method using a simulation study and real application on gas chromatography on Fast Odor Chromatographic Sniffer (FOX) system. / Doctor of Philosophy / A variety of complex data has broadened in many research fields such as epidemiology, genomics, and analytical chemistry with the development of science, technologies, and design scheme over the past few decades. For example, in epidemiology, the matched case-crossover study design is used to investigate the association between the clustered binary outcomes of disease and a measurement error in covariate within a certain period by stratifying subjects' conditions. In genomics, high-correlated and high-dimensional(HCHD) data are required to identify important genes and their interaction effect over diseases. In analytical chemistry, multiple time series data are generated to recognize the complex patterns among multiple classes. Due to the great diversity, we encounter three problems in analyzing the following three types of data: (1) matched case-crossover data, (2) HCHD data, and (3) Time-series data. We contribute to the development of statistical methods to deal with such complex data. First, under the matched study, we discuss an idea about hypothesis testing to effectively determine the association between observed factors and risk of interested disease. Because, in practice, we do not know the specific form of the association, it might be challenging to set a specific alternative hypothesis. By reflecting the reality, we consider the possibility that some observations are measured with errors. By considering these measurement errors, we develop a testing procedure under the matched case-crossover framework. This testing procedure has the flexibility to make inferences on various hypothesis settings. Second, we consider the data where the number of variables is very large compared to the sample size, and the variables are correlated to each other. In this case, our goal is to identify important variables for outcome among a large amount of the variables and build their network. For example, identifying few genes among whole genomics associated with diabetes can be used to develop biomarkers. By our proposed approach in the second project, we can identify differentially expressed and important genes and their network structure with consideration for the outcome. Lastly, we consider the scenario of changing patterns of interest over time with application to gas chromatography. We propose an efficient detection method to effectively distinguish the patterns of multi-level subjects in time-trend analysis. We suggest that our proposed method can give precious information on efficient search for the distinguishable patterns so as to reduce the burden of examining all observations in the data.

Cette thèse porte sur l'étude de modèles de polymère en milieu aléatoire: on se concentre sur le cas d'un polymère dirigé en dimension d+1 qui interagit avec un défaut unidimensionnel. Les interactions sont possiblement non-homogènes, et sont représentées par des variables aléatoires. Une question importante est celle de l'influence du désordre sur le phénomène de localisation: on veut déterminer si la présence d'inhomogénéités modifie les propriétés critiques du système, et notamment les caractéristiques de la transition de phase (auquel cas le désodre est dit pertinent). En particulier, nous prouvons que dans le cas où le défaut est une marche aléatoire, le désordre est pertinent en dimension d≥3. Ensuite, nous étudions le modèle d'accrochage sur une ligne de défauts possédant des inhomogénéités corrélées spatialement. Il existe un critère non rigoureux (dû à Weinrib et Halperin), que l'on applique à notre modèle, et qui prédit si le désordre est pertinent ou non en fonction de l'exposant critique du système homogène, noté νpur, et de l'exposant de décroissance des corrélations. Si le désordre est gaussien et les corrélations sommables, nous montrons la validité du critère de Weinrib-Halperin: nous le prouvons dans la version hiérarchique du modèle, et aussi, de manière partielle, dans le cadre (standard) non-hiérarchique. Nous avons de plus obtenu un résultat surprenant: lorsque les corrélations sont suffisamment fortes, et en particulier si elles sont non-sommables (dans le cadre gaussien), il apparaît un régime où le désordre devient toujours pertinent, l'ordre de la transition de phase étant toujours plus grand que νpur. La prédiction de Weinrib-Halperin ne s'applique alors pas à notre modèle. / This thesis studies models of polymers in random environment: we focus on the case of a directed polymer in dimension d+1 that interacts with a one-dimensional defect. The interactions are possibly inhomogeneous, and are represented by random variables. We deal with the question of the influence of disorder on the localization phenomenon: we want to determine if the presence of inhomogeneities modifies the critical properties of the system, and especially the characteristics of the phase transition (in that case disorder is said to be pertinent). In particular, we prove that if the defect is a random walk, disorder is relevant in dimension d≥3. We then study the pinning model in random correlated environment. There is a non-rigourous criterion (due to Weinrib and Halperin), that we can apply to our model, and that predicts disorder relevance/irrelevance, according to the value of the critical exponent of the homogeneous system, denoted νpur, and of the correlation decay exponent. When disorder is Gaussian and correlations are summable, we show that the Weinrib-Halperin criterion is valid: we prove this in the hierarchical version of the model, and also, partially, in the non-hierachical (standard) framework. Moreover, we obtained a surprising result: when correlations are sufficiently strong, and in particular when they are non-summable (in the gaussian framework), a new regime in which disorder is always relevant appears, the order of the phase transition being always larger than νpur. The Weinrib-Halperin prediction therefore does not apply to our model.

La SSD (Species Sensitivity Distribution) est une méthode utilisée par les scientifiques et les régulateurs de tous les pays pour fixer la concentration sans danger de divers contaminants sources de stress pour l'environnement. Bien que fort répandue, cette approche souffre de diverses faiblesses sur le plan méthodologique, notamment parce qu'elle repose sur une utilisation partielle des données expérimentales. Cette thèse revisite la SSD actuelle en tentant de pallier ce défaut. Dans une première partie, nous présentons une méthodologie pour la prise en compte des données censurées dans la SSD et un outil web permettant d'appliquer cette méthode simplement. Dans une deuxième partie, nous proposons de modéliser l'ensemble de l'information présente dans les données expérimentales pour décrire la réponse d'une communauté exposée à un contaminant. A cet effet, nous développons une approche hiérarchique dans un paradigme bayésien. A partir d'un jeu de données décrivant l'effet de pesticides sur la croissance de diatomées, nous montrons l'intérêt de la méthode dans le cadre de l'appréciation des risques, de par sa prise en compte de la variabilité et de l'incertitude. Dans une troisième partie, nous proposons d'étendre cette approche hiérarchique pour la prise en compte de la dimension temporelle de la réponse. L'objectif de ce développement est d'affranchir autant que possible l'appréciation des risques de sa dépendance à la date de la dernière observation afin d'arriver à une description fine de son évolution et permettre une extrapolation. Cette approche est mise en œuvre à partir d'un modèle toxico-dynamique pour décrire des données d'effet de la salinité sur la survie d'espèces d'eau douce / Species Sensitivity Distribution (SSD) is a method used by scientists and regulators from all over the world to determine the safe concentration for various contaminants stressing the environment. Although ubiquitous, this approach suffers from numerous methodological flaws, notably because it is based on incomplete use of experimental data. This thesis revisits classical SSD, attempting to overcome this shortcoming. First, we present a methodology to include censored data in SSD with a web-tool to apply it easily. Second, we propose to model all the information present in the experimental data to describe the response of a community exposed to a contaminant. To this aim, we develop a hierarchical model within a Bayesian framework. On a dataset describing the effect of pesticides on diatom growth, we illustrate how this method, accounting for variability as well as uncertainty, provides benefits to risk assessment. Third, we extend this hierarchical approach to include the temporal dimension of the community response. The objective of that development is to remove the dependence of risk assessment on the date of the last experimental observation in order to build a precise description of its time evolution and to extrapolate to longer times. This approach is build on a toxico-dynamic model and illustrated on a dataset describing the salinity tolerance of freshwater species

Neste trabalho estudamos a existência de ordem de longo alcance em modelos ferromagnéticos na presença de um campo externo cuja configuração apresenta um padrão tipicamente aleatório. Provamos por meio do argumento de Peierls modificado por Griffiths para o estudo de um antiferromagneto, que o modelo de Ising ferromagnético bidimensional exibe, para um campo alternado de intensidade fraca, ordem de longo alcance `a temperatura finita. Propomos dar um passo além considerando campos auto-similares esparsos, cuja soma é nula em todas as escalas. Estudamos também o modelo hierárquico em duas dimensões, para o qual provamos a existência de ordem de longo alcance a temperatura finita, na ausência de campo externo e para um campo com regiões irregulares esparsas. Provamos que os resultados do modelo de contornos hierárquicos são equivalentes aos resultados do modelo hierárquico em duas dimensões. Por fim, provamos através do método do limite infravermelho existência de ordem de longo alcance no modelo N-vetorial com campo alternado, de intensidade fraca, para d >= 3, sob a hipótese de que a variância do estado associado `a interação com o campo apresenta cardinalidade inferior a do volume do sistema. Mostramos, sob hipóteses similares, que o modelo N-vetorial hierárquico com campo externo, esparso e de intensidade pequena, apresenta ordem de longo alcance a baixas temperaturas. / In this work we study the existence of long range order for ferromagnetic models in the presence of an external field whose configuration has a pattern typically random. We prove, via the Peierls\' argument modified by Griffiths in his study of an antiferromagnet, that the two dimensional ferromagnetic Ising model for a staggered field exhibits long-range order at finite temperature and small field intensity. We propose to give a further step considering sparse self similar fields, whose sum is zero in all scales. We study as well the hierarchical model in two dimensions, where we prove existence of long-range order at finite temperature in the absence of external field and for a field configuration with sparse irregular regions. We prove that the results for the two-dimensional hierarchical contours model are equivalent to the results of the hierarchical model in two dimensions. Lastly, we prove via infrared bound method, existence of long range order in the N-vector model with a staggered and weak external field for d >= 3, under the hypothesis that the variance of the state connected with the field interaction has cardinality lower than volume. We show, under similar hypotheses, that the N-vector hierarchical model with a sparse field of low intensity has long range ordem at low temperatures.

Gelman, King, 及Liu(1998)針對一連串且互相獨立的橫斷面調查提出多重設算程序，且對不同調查的參數以階層模式(hierarchical model)連結。本文為介紹複雜抽樣(分層或群集抽樣)之下，若Q個連續變數有遺漏現象時，如何結合對象之個別特性，各層或各群集的參數，以及連結各層或各群集參數的階層模式，以設算遺漏值及估計模式中之參數。 對遺漏值的處理採用單調資料擴展演算法，只需對破壞單調資料型態的遺漏值進行設算。由於考慮到不同的群集或層往往呈現不同的特性，因而以階層模式連絡各群集或各層的參數，並將Gelman, King, Liu(1998)的推導結果擴展到將個別對象之特性納入考量之上。對各群集而言，他們的共變異數矩陣Ψ及Σ為影響群內其他參數的收斂情形，由模擬獲得的結果，沒有證據顯示應懷疑收斂的問題。 / Gelman, king, and Liu (1998) use multiple imputation for a series of cross section survey, and link the parameter of different survey by hierarchical model. This text introduces a method to impute missing value and estimate the parameters affected by hierarchical model if Q continuous variables has missing value under complex survey. For each cluster, the parameters are influenced by their variance-covariance matrix Ψ and Σ. The result obtained from the simulation have no clear evidence to doubt the convergence of parameters.