One of the fundamental tasks in science is to find explainable relationships between observed phenomena. Recent work has addressed this problem by attempting to learn the structure of graphical models - especially Gaussian models - by the imposition of sparsity constraints. The graphical lasso is a popular method for learning the structure of a Gaussian model. It uses regularisation to impose sparsity. In real-world problems, there may be latent variables that confound the relationships between the observed variables. Ignoring these latents, and imposing sparsity in the space of the visibles, may lead to the pruning of important structural relationships. We address this problem by introducing an expectation maximisation (EM) method for learning a Gaussian model that is sparse in the joint space of visible and latent variables. By extending this to a conditional mixture, we introduce multiple structures, and allow side information to be used to predict which structure is most appropriate for each data point. Finally, we handle non-Gaussian data by extending each sparse latent Gaussian to a Gaussian copula. We train these models on a financial data set; we find the structures to be interpretable, and the new models to perform better than their existing competitors. A potential problem with the mixture model is that it does not require the structure to persist in time, whereas this may be expected in practice. So we construct an input-output HMM with sparse Gaussian emissions. But the main result is that, provided the side information is rich enough, the temporal component of the model provides little benefit, and reduces efficiency considerably. The GWishart distribution may be used as the basis for a Bayesian approach to learning a sparse Gaussian. However, sampling from this distribution often limits the efficiency of inference in these models. We make a small change to the state-of-the-art block Gibbs sampler to improve its efficiency. We then introduce a Hamiltonian Monte Carlo sampler that is much more efficient than block Gibbs, especially in high dimensions. We use these samplers to compare a Bayesian approach to learning a sparse Gaussian with the (non-Bayesian) graphical lasso. We find that, even when limited to the same time budget, the Bayesian method can perform better. In summary, this thesis introduces practically useful advances in structure learning for Gaussian graphical models and their extensions. The contributions include the addition of latent variables, a non-Gaussian extension, (temporal) conditional mixtures, and methods for efficient inference in a Bayesian formulation.

Seismic hazard analysis aims to estimate human risk due to natural disasters such as earthquakes. To improve seismic hazard analysis, our group is focused on earthquake induced tsunamis and the use of statistical models to reconstruct historical earthquakes. Based on the estimated wave heights given in anecdotal historical descriptions, we created observational probability distributions to model the historically recorded observations and constructed a prior distribution on the relevant earthquake parameters based on known seismicity of a given region. Then we used the software package GeoClaw, and a Metropolis-Hastings sampler to obtain a posterior distribution of earthquake parameters that most closely matches the historically recorded tsunami. Our method was tested on two main events that occurred in 1820 and 1852 in central and eastern Indonesia respectively. The random walk Metropolis-Hastings sampler we employed appeared to recover the causal earthquake quite well, but the computational costs were prohibitive even though both scenarios we considered were relatively simple. To improve the sampling procedure, we have focused on advanced sampling techniques such as Hamiltonian Monte Carlo (HMC) where the gradient of the forward model (Geoclaw) is required. This is problematic however as this gradient is not available computationally. To mitigate this problem, we make use of a linearized adjoint solver for the shallow water equations, and exact gradient calculations for the Okada earthquake rupture model, yielding a surrogate gradient that leads to improved sampling.

TET2 et P53, deux suppresseurs de tumeurs, jouent un rôle important dans l’homéostasie des cellules souches hématopoïétiques et sont trouvés mutés dans les hémopathies malignes. Ils sont aussi impliqués dans le contrôle du cycle cellulaire et les mécanismes de réparation des dommages de l’ADN, notamment la voie de réparation par excision de base (BER). Dans la première partie de ce travail, nous avons montré que la surexpression de TET2 et l’augmentation consécutive des 5hmC, ralentit la progression du cycle cellulaire et la transition G1/S et induit une instabilité centrosomique associée à une instabilité chromosomique dans un modèle cellulaire Ba/F3. De plus, la surexpression de TET2 induit l’augmentation de la mutagenèse particulièrement des transitions C->T dans les sites CpG dans un contexte déficient en thymidine DNA glycosylase (TDG), une protéine initiatrice du BER. Dans la seconde partie de ce travail, nous avons montré que l’activation de P53, par des antagonistes de MDM2, a un effet délétère sur tous les progéniteurs hématopoïétiques. Ces antagonistes induisent aussi une cytotoxicité non seulement dans les stades précoces de la mégacaryopoïèse mais surtout dans les stades tardifs. Cette cytotoxicité n’est pas réversible, contrairement à ce qui est observé en clinique, et ne peut pas être restaurée par des doses croissantes de thrombopoïétine. Au total, TET2 et P53 doivent être strictement régulés pour assurer l’homéostasie et la stabilité génétique des cellules hématopoïétiques. / Two tumor suppresors, TET2 and P53, play an important role in the homeostasis of hematopoietic stem cells and have been found mutated in hematological malignancies. They are also involved in cell cycle control and DNA repair mechanisms, including the base excision repair pathway (BER). In the first part of this work, we showed that TET2 overexpression and the consequent increase of 5hmC, inhibit cell cycle progression particularly G1/S transition and induces centrosome instability associated with chromosomal instability in Ba/F3 cellular model. In addition, overexpression of TET2 induces increased mutagenesis particularly transitions C->T at CpG sites in a context deficient in thymidine DNA glycosylase (TDG), a protein initiating BER. In the second part of this work, we have shown that p53 activation by MDM2 antagonists has deleterious effect on all haematopoietic progenitors. These antagonists also induce cytotoxicity not only in the early stages of megakaryopoiesis but also mainly in the late stages. This cytotoxicity is not reversible, in contrast to what is observed in clinic, and can not be restored by increasing doses thrombopoietin. To conclude, TET2 and P53 must be strictly controlled to ensure homeostasis and genetic stability of the hematopoietic cells.

Although Set-top-Boxes(STBs) are widely deployed today to connect a media source to a display (traditionally a television set), the market is changing due to the introduction of Internet Protocol Television, Over-the-Top streaming devices, gaming console, home theater Personal Computer, smart TV, etc. There is an evolving concept of a Home Media Center (HMC). This HMC provides consumers with an integrated home media environment and experience. This thesis explores the transition from STBs to HMCs. The specific questions that this thesis project answers are: What will a future HMC look like? What will its functions be? What interfaces and protocols will it use? Who will make these HMCs?  How can STB vendors evolve to be HMC vendors or will they simply cease to exist?  This thesis project designed and evaluated a hypothetical HMC prototype based upon current technology trends and user expectations. This prototype was used with 68 volunteers to identify and prioritize the most important features that a HMC should provide. Based upon the most important of these features a conceptual HMC prototype is designed to define a HMC product roadmap for 1, 3, and 5 years. This roadmap is used to project the economic impact of HMCs on the current STB industry.  This economic analysis considers Sweden as the target market. This thesis could be used by current STB vendors to define their own company specific roadmaps to support their transition to the future HMC market. / Aven digitalboxar är spridda i dag för att ansluta en mediakälla till en bildskärm (traditionellt en TV) är marknaden förändras på grund av införandet av Internet Protocol Television, Over-the-Top streaming anordningar, spelkonsol, hem teater Personal Computer, smarta TV osv. Det finns en framväxande begreppet Home Media Center (HMC). Detta HMC ger konsumenterna ett integrerat hem mediemiljö och erfarenhet. Denna avhandling utforskar övergången från digitalboxar till HMC. De specifika frågor som detta examensarbete skall besvara är: Vad kommer en framtida HMC se ut? Hur kommer dess funktioner att bli? Vilka gränssnitt och protokoll kommer den att använda? Vem kommer att göra dessa HMC? Hur kan digitalboxar leverantörer utvecklas vara HMC säljare eller kommer de helt enkelt att upphöra att existera? Detta examensarbete designar och utvärderar en hypotetisk HMC prototyp baserad på nuvarande tekniktrender som användarnas förväntningar studie. Denna prototyp användes med 68 frivilliga att identifiera och prioritera de viktigaste funktionerna som en HMC bör ge. Baserat på de viktigaste av dessa funktioner en konceptuell HMC prototyp kommer att utformas för att definiera en HMC produkt färdplan för en, 3 och 5 år. Färdplanen kommer att användas för att projicera de ekonomiska konsekvenserna av HMC på den aktuella digitalboxar industrin. Denna ekonomiska analysen kommer att överväga Sverige som målgrupp. Denna avhandling kan användas av nuvarande digitalboxar leverantörer för att definiera sina egna företagsspecifika färdplaner för att stödja övergången till den framtida HMC marknaden.

科学研究費補助金 研究種目:一般研究(B)(2) 課題番号:07457240 研究代表者:宮田 敏男 研究期間:1995-1996年度

Este trabalho apresenta um estudo através da abordagem Bayesiana em modelos de volatilidade estocástica, para modelagem de séries temporais financeiras, com o uso do método de Monte Carlo Hamiltoniano (HMC). Propomos o uso de outras distribuições para os erros da equação de observações do modelos de volatilidade estocástica, além da distribuição Gaussiana, para tratar problemas como caudas pesadas e assimetria nos retornos. Além disso, utilizamos critérios de informações, recentemente desenvolvidos, WAIC e LOO que aproximam a metodologia de validação cruzada, para realizar a seleção de modelos. No decorrer do trabalho, estudamos a qualidade do método HMC através de exemplos, estudo de simulação e aplicação a conjunto de dados. Adicionalmente, avaliamos a performance dos modelos e métodos propostos através do cálculo de estimativas para o Valor em Risco (VaR) para múltiplos horizontes de tempo. / This paper presents a study using Bayesian approach in stochastic volatility models for modeling financial time series, using Hamiltonian Monte Carlo methods (HMC). We propose the use of other distributions for the errors of the equation at stochastic volatiliy model, besides the Gaussian distribution, to treat the problem as heavy tails and asymmetry in the returns. Moreover, we use recently developed information criteria WAIC and LOO that approximate the crossvalidation methodology, to perform the selection of models. Throughout this work, we study the quality of the HMC methods through examples, simulation study and application to dataset. In addition, we evaluated the performance of the proposed models and methods by calculating estimates for Value at Risk (VaR) for multiple time horizons.

Mammalian DNA methylathion is a chemical reaction catalyzed by DNA methyltransferases (DNMTs) and involves the addition of a methyl group from the methyl donor SAM to the carbon 5 position of cytosine (C) in a CpG dinucleotide. Specifically, DNA methylation is essential for normal development and is involved in numerous key mechanisms such as genomic imprinting, X-chromosome inactivation, suppression of repetitive elements and may be involved in the regulation of single-copy gene expression. In the human genome the majority of CpGs are methylated whereas regions with high density of CpG sites, termed CpG islands and often co-localized within gene promoters, are typically free of this mark. Recently, a new modified cytosine, 5-hydroxymhetylcytosine (5-hmC), was identified and found at significant levels in mouse brain and both mouse and human embryonic stem (ES) cells. The conversion of 5-mC to 5-hmC is catalyzed by the ten-eleven translocation (TET) proteins of the 2-oxoglutarate (2OG)-and Fe(II)-dependent oxygenase superfamily. Many studies were conducted since the identification of 5-hmC and significant levels of 5-mC hydroxylation were found in many other mouse and human tissues. Importantly, many of the techniques used for 5-mC detection, such as bisulphite sequencing and methyl-sensitive restriction digestion, are incapable of distinguishing between 5mC and 5hmC implying the necessity not only to develop techniques specific for 5-hmC characterization but also reevaluation of previously published 5mC data. The biological function of 5-hmC is unknown however many recent studies have suggested a role for 5-hmC as an intermediate of either passive or active demethylation. The majority of studies of 5- hmC and TETs have used mouse ES cells as model system. Therefore, very little is known about 5-hmC patterns and TET expression within and between normal tissues. During my PhD, I used the recently developed 5-hmC-specific antibody for tiling microarrays and 5hmC-qPCR to examine both global 5hmC content and locus-specific patterns of 5hmC in several normal human tissues and breast cancer. I found that global 5-hmC content is highly variable between tissues compared to global 5-mC content. Moreover, TETs genes are highly expressed in most of tissues tested. Importantly, both global 5-hmC content and TETs genes are rapidly and significantly reduced as consequence of adaptation of cells from normal human tissue to cell culture. Using the 5hmC-specific antibody for tiling microarrays and 5-hmC-qPCR to profile locus-specific patterns of 5hmC, I found that 5-hmC patterns are tissue-specific in human samples. In addition, comparing array data to RNA-seq data, 5- hmC was found to co-localize at gene bodies of active genes. Moreover, despite the global 5-hmC reduction in cell lines, 5-hmC content remains enriched in some specific loci. In summary, my results show that tissue type is a major modifier of both global and locus-specific 5hmC at genes in normal human tissues. Furthermore, I also show that both TET gene expression and 5hmC content are significantly reduced and 5-hmC profiles reprogrammed during the passage from tissues to cell culture.

La modélisation du comportement hydromécanique chimique des milieux poreux saturés et non saturés est abordée au niveau microscopique et mésoscopique. Au niveau microscopique la modélisation des écoulements diphasiques est basée sur une représentation du réseau poral comme un ensemble de tubes dont les orientations et les rayons sont choisis sur un principe d’équivalence avec les pores. L’algorithme régissant la génération des conduits et les écoulements d'eau et de gaz est développé en utilisant un langage propre au code 3FLO. En utilisant cette approche directe on étudie ensuite le problème de l’impact de l’endommagement sur les propriétés effectives hydriques en milieu saturé et non saturé. Sans expliciter l’origine et les mécanismes de la naissance et de la propagation des fissures, on simule le développement d’une fissuration au cours d’un essai de compression triaxiale en générant numériquement une fissuration orientée préférentiellement dans la direction de la contrainte maximale en compression. Les simulations réalisées en milieu saturé et non saturé mettent en évidence l’impact de la fissuration sur la conductivité hydraulique, la courbe de rétention et la perméabilité relative.Une dernière partie de la thèse est consacrée à la modélisation mésoscopique des couplages hydro-chimio-mécaniques par une approche de couplage indirecte ou dite de communication-passerelle. Selon cette méthodologie plusieurs codes spécialisés dans un domaine donné communiquent entre eux et sont capables de réaliser des calculs paramétrés par des champs externes. Un exemple est décrit ici par le couplage d’un code hydro-géochimie (HP1) et d'un code de calculs mécaniques développé sous Matlab. En utilisant cette méthodologie on traite d’abord le problème d’évolution du potentiel de gonflement d’une argile gonflante au cours d’une infiltration par une plume alcaline. Les simulations numériques mettent en évidence que pour la période étudiée la variation de la composition chimique est le mécanisme dominant en comparaison avec la dissolution des minéraux argileux. L’adaptation du modèle ELASGONF pour réaliser des calculs mécaniques paramétrés par des champs de concentration a permis de simuler la diminution de la pression de gonflement lors d’infiltration. En utilisant cette même approche, l’hypothèse de formation du gypse dans le tuffeau blanc par voie aérienne a été étudiée dans le cadre de la problématique de préservation des monuments historiques tels que les châteaux de la Loire. Dans les conditions de nos simulations, la formation du gypse par cette voie n’a pas été vérifiée. / Modelling of the behavior and the couplings HMC of the porous circles

No description available.

International Telemetering Conference Proceedings / October 30-November 02, 1995 / Riviera Hotel, Las Vegas, Nevada / The Remote Unit (RU) for a decentralised on-board base-band telemetry system is designed for use in launch vehicle missions of the Indian Space Research Organisation (ISRO). This new design is a highly improved and miniaturised version of an earlier design. The major design highlights are as follows. Usage of CMOS Field Programmable Gate Array (FPGA) technology in place of LS TTL devices, the ability to acquire various types of data like high level single ended or differential analog, bi-level events and two channels of high speed asynchronous serial data from On-Board Computers (OBCs), usage of HMC technology for the reduction of discrete parts etc. The entire system is realised on a single 6 layer MLB and is packaged on a stackable modular frame. This paper discusses the design approach, tools used, simulations carried out, implementation details and the results of detailed qualification tests done on the realised qualification model.