• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 527
  • 262
  • 220
  • 32
  • 27
  • 19
  • 17
  • 13
  • 10
  • 9
  • 8
  • 8
  • 8
  • 5
  • 4
  • Tagged with
  • 1311
  • 181
  • 154
  • 150
  • 121
  • 113
  • 112
  • 101
  • 100
  • 93
  • 87
  • 85
  • 84
  • 81
  • 75
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
51

Redução no vício da distribuição da deviance para dados de contagem. / Bias reduction in the distribution of the deviance for count data.

Denise Nunes Viola 26 October 2001 (has links)
Dados de contagem podem ser considerados, em geral, como provenientes de uma distribuição de Poisson. Neste contexto, a análise de tais dados apresenta certas dificuldades, pois não segue algumas pressuposições básicas para o ajuste de um modelo matemático. Desse modo, algumas transformações são sugeridas, mas nem sempre bons resultados são obtidos. No enfoque de Modelos Lineares Generalizados, a estatística que mede a qualidade do ajuste do modelo para os dados é chamada deviance. Porém, a distribuição da deviance é, em geral, desconhecida. No entanto, para dados com distribuição de Poisson, pode-se mostrar que a distribuição da deviance se aproxima de uma distribuição ?2, mas tal aproximação não é boa para tamanhos pequenos de amostra. Para melhorar essa aproximação, alguns fatores de correção para os dados são sugeridos, mas os resultados obtidos ainda não são satisfatórios. Assim, o objetivo deste trabalho é propor um novo fator de correção para os dados seguindo uma distribuição de Poisson, de modo a se obter uma melhora na distribuição da deviance para qualquer tamanho de amostra. Para isto, será adicionada uma constante à variável resposta e, através do valor esperado da deviance, calcula-se tal constante de modo a reduzir o erro cometido na aproximação. Para verificar a melhora na aproximação da distribuição da deviance a uma distribuição qui-quadrado, dados de uma distribuição de Poisson são simulados e o valor da deviance é calculado. QQ-plots são construídos para a comparação com a distribuição qui-quadrado. / Analysis of count data presents, in general, can be supposed coming from a Poisson distribution. The analysis of such data have some problems once the underlying distribution of them does not follow the basic assumptions to fit a model. Some tranformations can be suggested, but good results are not always obtained. In the approach of the Generalized Linear Models, the deviance is the statistics that measures the goodness of fit, but its distribution is unknown. Furthermore, considering Poisson distribution data, it is possible to approximate the distribution of the deviance for a chi-square distribution, but such approximation is not good for small sample size. In order of improve this approximation, corrections for the data are suggested, but the results are not good yet. Then, the aim of this work is to propose a new correction factor for data following a Poisson distribution in order to obtain an improvement in the distribution of the deviance for any sample size. For this, just adding a constant at the response variable and, through the expected value of the deviance, such constant is obtained in order to reduce the error in the aproximation. Simulated data from the Poisson distribution were made to calculate the deviance with and without the correction and QQ-plots were used to compare the values of the deviance with the chi-square distribution.
52

Modelagem para dados de parasitismo / Modelling parasitism data

João Mauricio Araújo Mota 23 September 2005 (has links)
Experimentos com diferentes objetivos têm sido conduzidos a fim de se estudar o mecanismo do parasitismo, sendo muito comuns os bioensaios para encontrar condições ótimas para a produção de parasitas e para definir estratégias para liberações inundativas no campo. Assim, por exemplo, o número de ovos parasitados depende de fatores como: espécie, tipo e densidade do hospedeiro, longevidade do adulto e densidade do parasita, tipo de alimentação, temperatura, umidade etc. Logo, o objetivo de um determinado ensaio pode ser, então, estudar o comportamento da variável resposta como função do número de parasitóides ou do número de hospedeiros ou ainda do tipo de alimentação. Pode-se verificar que, em geral, as variáveis observáveis so contagens ou somas aleatórias de variáveis aleatórias ou proporções com denominadores fixos ou aleatórios. A distribuição padrão para modelar contagens é a Poisson enquanto que para proporções é a binomial. Em geral, elas não se ajustam dados oriundos do processo de parasitismo, pois suas pressuposições não são satisfeitas, e surgiram modelos alternativos e que levam em consideração o mecanismo de evitar o superparasitismo. Alguns deles supõem que a probabilidade de fuga (evitar o superparasitismo) é função do número de ovos presentes no hospedeiro (Bakker, 1967,1972; Rogers, 1975; Griffits, 1977), outros não consideram tal processo (Daley; Maindonald, 1989; Griffths, 1977). Outros, ainda, incluem o comportamento seletivo do parasita na escolha do hospedeiro e a habilidade do hospedeiro em atrair o parasita (Hemerik et al, 2002). Alguns deles surgiram independemente, outros como generalizações, sendo, portanto, de interesse um estudo adicional para ressaltar pontos comuns entre eles. No presente trabalho, são estudados 19 modelos probabilísticos para explicar a distribuição do número de ovos postos por um parasita em um determinado hospedeiro. Foi mostrada a equivalência entre alguns deles e, além disso, foi provado que um modelo usado por Faddy (1997) na estimação do tamanho de população animal generaliza-os. As propriedades desse modelo são apresentadas e discutidas. O uso do modelo de Faddy para a distribuição do número de ovos no sistema parasita-hospedeiro é o principal resultado teórico dessa tese. / Experiments with distinct aims have been conducted in order to study the parasitism mechanism. Bioassays to find optimum conditions for parasite production and to define strategies for application in the field are very common. In this way, for example, the number of parasitized eggs depends on factors such as: species, type and host density, adult longevity and parasite density, type of food, temperature, humidity etc. Hence, the objective of a specific assay can be to study the response variable behavior as a function of the number of parasitoids, number of hosts or type of food. In general, the observable variables are counts or random sums of random variables or proportions with fixed or random denominators. Poisson is the standard distribution used to model counts, while the distribution used for proportions is the binomial. In general, they do not fit to standard distributions don’t fit to data generated by the parasitism process, because their assumptions are not satisfied and alternative models that consider the mechanism of avoidance of superparasitism have appeared in the literature. Some of them consider that the refuse probability (avoidance of superparasitism) is a function of the number of eggs in the host (Bakker, 1967, 1972; Rogers, 1975; Griffiths, 1977) others don’t consider such process (Daley and Maindonald 1989; Griffiths, 1977). Some include the selective behavior of the parasite in the choice of the host and the host ability in attract the parasite (Hemerick et al., 2002). Some of the models have appeared independently, others as generalizations. There is therefore some interest in making a study of the common points of the models. In this work 19 probability models found in the literature are presented to explain the distribution of the number of eggs laid by a parasite on a specific host. As an initial result, the equivalence between some of these models is shown. Also it is shown that a model developed by Faddy (1997) can be considered as a generalization of 18 of the models. The properties of this model are presented and discussed. The equivalence between the Janardan and Faddy models is an original and interesting result. The use of Faddy´s model as a general probability model for the distribution of the number of eggs in a parasite-host system is the main theoretical result of this thesis.
53

Simulação estocástica de variáveis aleatórias Poisson correlacionadas: aplicação ao controle populacional do percevejo (Euschistus heros Fabricius) da soja (Glycine max L.) / Stochastic simulation for correlated Poisson random variables: application to population control bedbug (Euschistus heros Fabricius) soy (Glycine max L.)

Raphael Antonio Prado Dias 07 March 2014 (has links)
A simulação de dados que seguem distribuição de Poisson é essencial em muitas aplicações reais de várias áreas, tais como saúde, marketing, ciências agronômicas, entre outras em que os dados são contagens multivariadas. Métodos de simulação atuais sofrem de limitações computacionais e restrições à estrutura de correlação e, portanto, são raramente usados. Neste trabalho propôs-se uma modificação do método NORTA para gerar dados com distribuição Poisson multivariada a partir de uma distribuição normal multivariada com matriz de correlações e vetor de médias pré estabelecidos. Como as distribuições Normal multivariada e univariada e a distribuição Poisson univariada já estão implementadas em softwares estatísticos, inclusive no R, implementou-se algumas linhas de código. Mostrou-se que o método funciona bem e é altamente preciso na geração de dados multivariados com distribuição marginais de Poisson, para diferentes estruturas de correlações (negativas e positivas e variando os valores) e para altos e baixos valores de médias. Mostrou-se as vantagens práticas da simulação de dados de Poisson multivariada sobre a normal multivariada na detecção da taxa de falsos alertas de super populações de percevejos, evidenciando que simulações inadequadas podem levar a excesso de falsos alertas. Uma vez que os dados seguem distribuição Poisson multivariada, a taxa de falsos alertas pode ser maior do que a imaginada. Essa taxa pode ser estimada por um modelo ajustado. A mesma técnica pode ser aplicada em diversos problemas de várias áreas do conhecimento. / The simulation data that follow a Poisson distribution is essential in many real applications in various areas such as healthcare, marketing, agronomic sciences, among others that the data are multivariate counts. Current simulation methods suffer from limitations and constraints on computing correlation structure and are therefore seldom used. This paper proposed a modification of the NORTA method for generating data with multivariate Poisson distribution from a multivariate normal distribution with correlation matrix and vector of predetermined average. As the multivariate and univariate Normal distribution and univariate Poisson distribution are already implemented in statistical software, including R, was implemented just a few lines of code. It was shown that the method works well and is highly accurate in generating multivariate data with marginal Poisson distribution structures for different correlations (negative and positive values) and for high and low ?. Proved the practical benefits of the simulation data on the multivariate Poisson multivariate normal in the detection of super bugs populations, inadequate simulations can lead to excessive false alerts. Once the data are multivariate Poisson distribution, the rate of false alarms can be greater than the imagined. This rate can be estimated by an adjusted model. The same technique can be applied to many problems in various fields of knowledge.
54

Topologie algébrique de complexes simpliciaux aléatoires et applications aux réseaux de capteurs / Algebraic topology of random simplicial complexes and applications to sensor networks

Ferraz, Eduardo 22 February 2012 (has links)
Cette thèse est composée de deux parties. La première partie utilise l’analyse stochastique pour fournir des bornes pour la probabilité de surcharge de différents systèmes grâce aux inégalités de concentration. Bien qu’ils soient généraux, nous appliquons ces résultats à des réseaux sans-fil réels tels que le WiMax et le traffic utilisateur multi-classe dans un système OFDMA. Dans la seconde partie, nous trouvons des liens entre la topologie de la couverture dans un réseau de capteur et celle du complexe simplicial correspondant. Cette analogie met en valeur de nouvelles facettes des certains objets mathématiques comme les nombres de Betti, le nombre de k-simplexes, et la caractéristique d’Euler. Puis, nous utilisons conjointement la topologie algébrique et l’analyse stochastique, en considérant que les positions des capteurs sont une réalisation d’un processus ponctuel de Poisson. Nous en déduisons les statistiques du nombre de k-simplexe et de la caractéristique d’Euler, ainsi que des bornes supérieures pour la distribution des nombres de Betti, le tout en d dimen- sions. Nous démontrons aussi que le nombre de k-simplexes converge vers une distribution Gaussienne quand la densité de capteurs tend vers l’infini à une vitesse de convergence connue. Enfin, nous nous limitons au cas unidimensionnel. Dans ce cas, le problème devient équivalent à résoudre une file M/M/1/1 préemptive. Nous obtenons ainsi des résultats analytiques pour des quantités telles que la distribution du nombre de composantes connexes et la probabilité de couverture totale. / This thesis has two main parts. Part I uses stochastic anlysis to provide bounds for the overload probability of different systems thanks to concentration inequalities. Although the results are general, we apply them to real wireless network systems such as WiMax and mutliclass user traffic in an OFDMA system. In part I I, we find more connections between the topology of the coverage of a sensor network and the topology of its corresponding simplicial complex. These connections highlight new aspects of Betti numbers, the number of k-simplices, and Euler characteristic. Then, we use algebraic topology in conjunction with stochastic analysis, after assuming that the positions of the sensors are points of a Point point process. As a consequence we obtain, in d dimensions, the statistics of the number of k-simplices and of Euler characteristic, as well as upper bounds for the distribution of Betti numbers. We also prove that the number of k-simplices tends to a Gaussian distribution as the density of sensors grows, and we specify the convergence rate. Finally, we restrict ourselves to one dimension. In this case, the problem becomes equivalent to solving a M/M/1/1 preemptive queue. We obtain analytical results for quantites such as the distribution of the number of connected components and the probability of complete coverage.
55

A familial longitudinal count data study

Goren, Hakan 14 October 2014 (has links)
In this report, I study familial longitudinal count data with a Poisson regression model. The data is collected from individuals who are nested in families. I focus on two main issues to fit a model. The first one is the large number of excess zeros and the second one is multi-level random effects. My approach for solving these problems are to use either Zero Inflated Poisson (ZIP) or Negative Binomial (NB) models to control for the excess zeros which allow for estimation of another parameter for over dispersion while developing the model with individual and familial random effects. First, I use a Poisson regression model with only main effects. After that, I fit a ZIP model to control for the extra zeros. I provide information about general form of the exponential families and a discussion about the dispersion parameter. I also fit a Negative Binomial model instead of the ZIP model. I also build these models with only individual random effects and with both individual and familial random effects as well. I discuss the generalized estimating equation (GEE) approach to estimate the parameters of a generalized linear model with auto regressive correlation between outcomes. / text
56

Etude du rôle du facteur de transcription Evi1 au cours du développement du rein embryonnaire chez le xénope

Van Campenhout, Claude C A 28 July 2006 (has links)
Chez les vertébrés, le système excréteur se développe de manière séquentielle sous la forme de trois types de reins différents : le pro-, le méso- et le métanéphros. Le pronéphros des embryons de xénope et de poisson zèbre se développe rapidement et présente une structure simple formée d’un unique néphron, ce qui en fait un excellent modèle d’étude de la néphrogenèse. Au cours du développement, le mésoderme pronéphrique est régionalisé en plusieurs domaines à l’origine des différents composants du néphron partagés par tous les reins des vertébrés : le glomus, les tubules proximaux, le tubule distal et le canal. Cette régionalisation fait appel à des mécanismes moléculaires encore peu connus. Parmi ceux-ci, le facteur de transcription WT1 inhiberait, dans les cellules à l’origine du glomus, l’expression des gènes caractéristiques des tubules proximaux. De plus, la voie de signalisation Notch est requise séquentiellement d’abord pour la formation du glomus et ensuite lors de la différenciation des tubules proximaux. Le gène Evi1 code pour un facteur de transcription à doigts à zinc notamment exprimé dans le métanéphros des vertébrés supérieurs et jouant un rôle majeur mais mal compris dans le développement. Chez l’embryon de xénope, le gène Evi1 est exprimé dès la fin de la neurulation dans la région ventro-postérieure de l’ébauche pronéphrique à l’origine du tubule distal et du canal. Son expression est inhibée par le facteur de transcription xWT1 et l’activation de la voie de signalisation Notch. Via des expériences de surproduction de la protéine Evi1 sauvage ou d’une fusion Evi1-VP16, fonctionnant de manière antagoniste à la protéine sauvage, nous avons montré que la protéine Evi1 joue un rôle important dans la néphrogenèse précoce en inhibant la formation du glomus et des tubules proximaux dans les cellules à l’origine du segment distal du néphron. Afin de déterminer l’importance du facteur de transcription xWT1 dans la régulation de l’expression du gène Evi1 ainsi que dans la formation du glomus, des expériences de sous-expression ont été réalisées. Ces expériences montrent que la sous-expression de xWT1 inhibe la formation du glomus mais n’induit cependant pas d’expansion de l’expression du gène Evi1 ni celle d’autres marqueurs du pronéphros, suggérant l’existence d’au moins un autre facteur répresseur exprimé au niveau de la couche médiane du mésoderme pronéphrique. Ensuite, nous avons comparé les profils de gènes codant notamment pour des protéines spécialisées dans le transport de solutés dans les reins embryonnaires de xénope et de poisson zèbre. Nos résultats montrent que le pronéphros du poisson zèbre, bien que présentant une structure uniforme, peut être subdivisé en quatre différents segments. Ces observations suggèrent que les reins embryonnaires du xénope et du poisson zèbre présentent des organisations similaires. Dans la dernière partie de ce travail, nous avons entamé l’étude du promoteur du gène CLC-K, codant pour un canal chlore exprimé dans du segment distal du néphron, par la réalisation d’embryons transgéniques de xénope. Les résultats préliminaires obtenus indiquent qu’un fragment de 11kb du promoteur du gène CLC-KB humain est suffisant pour diriger une expression correcte du transgène chez le xénope.
57

Theoretical and numerical treatment of singularities in elliptic boundary value problems

Beagles, A. E. January 1987 (has links)
No description available.
58

Performance of multi-state Markov modulated queuing in ATM networks

Yousef, Sufian Yacoub Salameh January 1998 (has links)
No description available.
59

Modelling and inference for a class of doubly stochastic point processes

Wei, Gang January 1995 (has links)
No description available.
60

Sample Size Determination in Auditing Accounts Receivable Using a Zero-Inflated Poisson Model

Pedersen, Kristen E 28 April 2010 (has links)
In the practice of auditing, a sample of accounts is chosen to verify if the accounts are materially misstated, as opposed to auditing all accounts; it would be too expensive to audit all acounts. This paper seeks to find a method for choosing a sample size of accounts that will give a more accurate estimate than the current methods for sample size determination that are currently being used. A review of methods to determine sample size will be investigated under both the frequentist and Bayesian settings, and then our method using the Zero-Inflated Poisson (ZIP) model will be introduced which explicitly considers zero versus non-zero errors. This model is favorable due to the excess zeros that are present in auditing data which the standard Poisson model does not account for, and this could easily be extended to data similar to accounting populations.

Page generated in 0.0361 seconds