Dissertations / Theses on the topic 'Análisis bayesiano'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 dissertations / theses for your research on the topic 'Análisis bayesiano.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
Cruz, Sarmiento Marylía Paola. "Análisis bayesiano de modelos de clases latentes para variables politómicas: Confianza hacia instituciones públicas." Master's thesis, Pontificia Universidad Católica del Perú, 2018. http://tesis.pucp.edu.pe/repositorio/handle/123456789/13457.
Full textTesis
Blas, Oyola Sthip Frank. "Métodos de selección de variables bajo el enfoque bayesiano para el modelo lineal normal." Master's thesis, Pontificia Universidad Católica del Perú, 2020. http://hdl.handle.net/20.500.12404/17868.
Full textTroncoso, Rojas Catalina Pía. "Determinación de precios óptimos para una cadena de supermercado utilizando modelos jerárquicos bayesianos." Tesis, Universidad de Chile, 2010. http://www.repositorio.uchile.cl/handle/2250/103749.
Full textParejas, Espinoza Héctor Eduardo. "Estimación de demanda con información incompleta para apoyar negociaciones de precios en una empresa industrial." Tesis, Universidad de Chile, 2012. http://www.repositorio.uchile.cl/handle/2250/111869.
Full textPara identificar oportunidades en el mercado es fundamental aumentar el conocimiento de los clientes y de la competencia. En el caso del marketing industrial, esto presenta un grado de dificultad mayor debido a que variables importantes como el precio final de venta y las ventas de la competencia son desconocidas. El presente trabajo se desarrolla para a un distribuidor de materias primas que vende a restaurantes. Desde la perspectiva del distribuidor es de interés entender cómo sus clientes deciden sus compras y optan por adquirir a un proveedor u otro. Para esto de gran utilidad poder inferir parte esta información no observable con la finalidad de mejorar sus políticas de precio. Como fuente de información se cuenta con datos transaccionales de venta de la empresa y una encuesta con características demográficas de los clientes. De esta forma surge como objetivo apoyar las negociaciones de precio del distribuidor mediante un mayor conocimiento del mercado en términos de su posición competitiva y un mayor entendimiento del comportamiento de compra de sus clientes. Para esto se desarrolla un modelo que considera el uso de la teoría económica para interpretar la elección de compra de los clientes. En la estimación de los parámetros del problema se utiliza un modelo lineal jerárquico que permita capturar heterogeneidad de los clientes identificando patrones de comportamiento. Dentro de los resultados obtenidos se logra identificar patrones de comportamiento de compra comunes en los clientes según sus características demográficas, lo cual sumado al modelo de decisión de compra hace posible inferir las compras de los clientes tanto al distribuidor como a su competencia, además de recomendaciones para los procesos de negociación. La estimación de las cantidades vendidas por el distribuidor de referencia bajo el efecto de la competencia arroja un error MAPE de 92,77% y un ajuste R2 de 0,6977. Finalmente dentro de las recomendaciones formuladas destaca que los precios de los clientes cuyo proveedor principal es el distribuidor de referencia son menos sensibles a compras realizadas en el pasado. Otro punto relevante es como la ubicación del cliente juega un papel relevante en prácticamente la totalidad de las variables explicativas del modelo, desde diferencias en los requerimientos básicos por categorías, variaciones en la sensibilidad al nivel de precios y las variables que describen las diferencias de precios.
Andrade, Chávez Francisco Mauricio. "Modelo de regresión Dirichlet bayesiano: aplicación para estimar la prevalencia del nivel de anemia infantil en centros poblados del Perú." Master's thesis, Pontificia Universidad Católica del Perú, 2020. http://hdl.handle.net/20.500.12404/18683.
Full textBernardo, Jose M. "Análisis de datos y métodos bayesianos." Pontificia Universidad Católica del Perú, 2013. http://repositorio.pucp.edu.pe/index/handle/123456789/95635.
Full textFilho, Paulo Cilas Marques. "Análise bayesiana de densidades aleatórias simples." Universidade de São Paulo, 2011. http://www.teses.usp.br/teses/disponiveis/45/45133/tde-25052012-184549/.
Full textWe define, from a known partition in subintervals of a bounded interval of the real line, a prior distribution over a class of densities with respect to Lebesgue measure constructing a random density whose realizations are nonnegative simple functions that integrate to one and have a constant value on each subinterval of the partition. These simple random densities are used in the Bayesian analysis of a set of absolutely continuous observables and the prior distribution is proved to be closed under sampling. We explore the prior and posterior distributions through stochastic simulations and find Bayesian solutions to the problem of density estimation. Simulations results show the asymptotic behavior of the posterior distribution as we increase the size of the analyzed data samples. When the partition is unknown, we propose a choice criterion based on the information contained in the sample. In spite of the fact that the expectation of a simple random density is always a discontinuous density, we get smooth estimates solving a decision problem where the states of nature are realizations of the simple random density and the actions are smooth densities of a suitable class.
Torres, Gil Santiago. "Simulación de Monte Carlo de la población de enanas blancas de la galaxia." Doctoral thesis, Universitat Politècnica de Catalunya, 2002. http://hdl.handle.net/10803/6578.
Full textBajo estas condiciones hemos desarrollado un modelo, basado en las técnicas de simulación Monte Carlo, de la población de enanas blancas, tanto del disco como del halo galáctico. Las simulaciones realizadas presentan los avances más recientes en la física de las enanas blancas a la par que incluyen de manera realista el proceso de selección observacional. Con la construcción de un modelo preciso podemos extraer la mayor información posible de los datos observacionales, analizar los posibles sesgos estadísticos derivados del proceso de selección, al igual que comprobar los resultados de los modelos teóricos o estimar futuras predicciones.
En primer lugar hemos realizado un estudio detallado y exhaustivo de la población de enanas blancas del disco galáctico. Este estudio se ha centrado en dos grandes bloques: el análisis de las propiedades cinemáticas y el estudio de la función de luminosidad. En este sentido hemos comprobado que los efectos de la ley de altura patrón no son en modo alguno despreciables y que pueden influir en la determinación de la edad del disco. Igualmente hemos analizado de forma especial la función de luminosidad, su completitud, sus sesgos observacionales, y una valoración estadística basada en las técnicas de estimación bayesiana de la edad del disco, obteniendo un valor de 13.5 Gyr con un error típico de 0.8 Gyr.
A continuación hemos analizado los posibles efectos de un episodio de mezcla en el disco galáctico. Tras analizar diferentes escenarios hemos podido comprobar que un tal episodio de mezcla tendría efectos nulos en la función de luminosidad, mientras que por el contrario existirían efectos apreciables en la distribución cinemática.
Por otra parte, mediante la implementación de una algoritmo de inteligencia artificial y utilizando nuestras simulaciones de la población del disco y del halo hemos realizado una clasificación de la población de enanas blancas a partir de un catálogo observacional. Esta clasificación nos ha permitido construir una función preliminar de las enanas blancas del halo que mejora substancialmente los datos anteriores.
Por último hemos analizado con detalle las propiedades de la población de enanas blancas del halo. En particular, hemos estudiado la contribución de estos objetos a la materia oscura de la galaxia. A tal efecto, hemos simulado los experimentos de microlentes en la dirección de la Gran Nube de Magallanes, al igual que estudiado el número de posibles objetos detectables por el Hubble Deep Field.
Vergara, Duarte Montserrat. "Causa de "Mortalidad prematura evitable". Estrategias de clasificación y aplicación al análisis de la desigualdad geogràfica en España." Doctoral thesis, Universitat Pompeu Fabra, 2009. http://hdl.handle.net/10803/7204.
Full textLos objetivos principales de esta tesis son: primero, proponer una nueva clasificación de causas de "mortalidad tratable"; segundo, analizar esta mortalidad en áreas pequeñas (municipios o agregados de municipios) de España; y tercero, mostrar la utilidad de dichos análisis en un formato accesible. La metodología usada en la tesis incluye una revisión de la literatura científica para valorar el nivel de eficacia de las intervenciones médicas disponibles para evitar la muerte por causas "tratables", un análisis empírico de dichas causas, una valoración de expertos, y la estimación del riesgo relativo de "mortalidad tratable" ajustado por edad mediante el uso de técnicas estadísticas bayesianas.
Esta tesis presenta una original propuesta de clasificación y análisis de causas de "mortalidad tratable" según el nivel de eficacia de las intervenciones médicas. Esta clasificación puede tener gran utilidad en la identificación de desigualdades geográficas y posibles deficiencias relacionadas con la efectividad de los servicios de salud.
"Avoidable mortality" is a relevant indicator to assess the quality of health care services. Studies on "avoidable mortality" distinguish between causes of death which are "preventable" (avoidable with public health interventions) and those which are "amenable" (avoidable with health services interventions). However, the two groups of causes are rather heterogeneous, making analysis and interpretation of results difficult.
The main aims of this thesis are: first, to propose a new classification of causes of "amenable mortality"; second, to analyse amenable mortality in small areas (municipalities or aggregations of municipalities) in Spain; and third, to show the usefulness of those analysis in an accessible format. Methodologies used in this thesis include a scientific literature review to assess the level of efficacy of available medical interventions to avoid death by "amenable" causes, an empirical analysis of those causes, an expert assessment and the estimation of age-adjusted relative risk of "amenable mortality" by Bayesian statistical techniques.
This thesis presents an original proposal for the classification and analysis of causes of "amenable mortality" according to the level of efficacy of medical interventions. This classification could be particularly useful in the identification of geographical inequalities and potential deficiencies related to the effectiveness of health care services
Gironès, Güell Xavier. "Metodología y análisis de la fabricación de anhidrita en horno rotativo mediante elementos de inteligencia artificial." Doctoral thesis, Universitat de Girona, 2013. http://hdl.handle.net/10803/101468.
Full textEn la fabricación de yeso en polvo, aplicable y controlable en su endurecimiento, se utilizan varios componentes. Uno de los componentes intrínseco del yeso para construcción es la anhidrita o yeso totalmente deshidratado que trabaja como parte inerte. Las redes bayesianas, como sistema experto, son una herramienta de gestión industrial. Una de sus propiedades más importantes es su capacidad de autoaprendizaje. Esta investigación se basará en la mejora del proceso de calcinación de anhidrita usando un horno rotativo de cocción directa. El trabajo aportará como novedad una actualización de los sistemas usados actualmente para la fabricación de anhidrita, ya que aparte de mejorar el control del proceso mediante lazos de control más eficientes y autogestionados, así como la introducción de mejoras en los niveles MES y Scada, aportará una modelización del proceso con elementos de inteligencia artificial mediante la aplicación de dichas redes bayesianas.
Pissini, Carla Fernanda. "Aplicações em meta-análise sob um enfoque bayesiano usando dados médicos." Universidade Federal de São Carlos, 2006. https://repositorio.ufscar.br/handle/ufscar/4593.
Full textFinanciadora de Estudos e Projetos
In this work, we consider the use of Meta-analysis with a Bayesian approach. Meta-analysis is a statistical technique that combines the results of di¤erent independent studies with purpose to find general conclusions. This term was introduced by Glass (1976) and it has been used when the number of studies about some research project is small. Usually, the models for Meta-analysis assume a large number of parameters and the Bayesian approach using MCMC (Markov Chain Monte Carlo) methods is a good alternative to combine information of independent studies, to obtain accutrate inferences about a specified treatment. As illustration, we consider real medical data sets on di¤erent studies, in which, we consider fixed and random e¤ects models. We also assume mixture of normal distributions for the error of the models. Another application is considered with educational data.
Neste trabalho, consideramos o uso de Meta-análise sob um enfoque Bayesiano. Meta-análise é uma técnica estatística que combina resultados de diversos estudos in-dependentes, com o propósito de descrever conclusões gerais. Este termo foi introduzido por Glass (1976) usado quando o número de estudos sobre alguma pesquisa científica é pequeno. Os modelos propostos para Meta-análise usualmente assumem muitos parâmetros e o enfoque Bayesiano com MCMC (Monte Carlo em Cadeias de Markov) é uma alternativa apropriada para combinar informações de estudos independentes. O uso de modelos Bayesianos hierárquicos permite combinações de vários estudos independentes, para a obtenção de inferências precisas sobre um determinado tratamento. Como ilustração numérica consideramos conjuntos de dados médicos de diferentes estudos e, na análise, utilizamos modelos de efeitos fixos e aleatórios e mistura de distribuições normais para o erro do modelo de regressão. Em uma outra aplicação relacionamos Meta-análise e Educação, através do efeito da espectativa do professor associada ao QI dos estudantes.
Sibim, Alessandra Cristiane. "Estimação e diagnóstico na distribuição exponencial por partes em análise de sobrevivência com fração de cura." Universidade de São Paulo, 2011. http://www.teses.usp.br/teses/disponiveis/55/55134/tde-09062011-151222/.
Full textThe main objective is to develop procedures inferences in a bayesian perspective for survival models with (or without) the cure rate based on piecewise exponential distribution. The methodology is based on bayesian methods for Markov Chain Monte Carlo (MCMC). To detect influential observations in the models considering bayesian case deletion influence diagnostics based on the Kullback-Leibler divergence (Cho et al., 2009). Furthermore, we propose the negative binomial model destructive cure rate. The proposed model is more general than the survival models with cure rate, since the probability to estimate the number of cases which were not eliminated by an initial treatment
Almeida, Julia Calhau. "Análise filogenética de Mydinae (Insecta, Diptera, Mydidae) com base em caracteres morfológicos e moleculares." Universidade de São Paulo, 2013. http://www.teses.usp.br/teses/disponiveis/41/41133/tde-22072013-153555/.
Full textThe Mydinae (Insecta, Diptera, Mydidae) occur only in the Americas and comprise 12 genera and 84 species, of which the vast majority of mydids occurring in Brazil belonging to this subfamily. Mydinae is currently divided into four tribes: Dolichogastrini, Messiasiini, Mydini and Phylomydini. The monophyly of the subfamily, as well as the monophyly of their tribes and genera, had not yet been tested by phylogenetic analysis. Concerning this fact, the objectives of this work are: 1) test the monophyly of the subfamily Mydinae, 2) check the phylogenetic relationship between Mydinae and other subfamilies of Mydidae, 3) test the monophyly of the tribes, subtribes and genera of Mydinae, as well as the monophyly of the species-groups of the genus Mydas; 4) propose a new classification of the subfamily based on phylogenetic results. The data from the external morphology of adults, and also DNA sequence of the COI gene, two methods of analysis were used: parsimony analysis with equal weighting of characters, and Bayesian probabilistic analysis. For each method, morphological and molecular data were analyzed separately and also in combination. The monophyly of Mydinae, as defined in the current classification, is not borne out in the present study. In both analyzes with morphological data, and Bayesian analysis with morphological and molecular data, a clade formed by all Mydinae (except Messiasia wilcoxi) + Paramydas (\'Apiophorinae\') was recovered. Among the tribes of Mydinae, the monophylies of Messiassiini and Mydini were not recovered. The genera Ceriomydas, Stratiomydas, Phyllomydas and Protomydas are recognized as natural groups. In the other hand, the genera Baliomydas, Gauromydas, Messiasia and Mydas did not form monophyletic groups in any of the conducted analyzes. Concerning the Mydas species-groups, only the interruptus group was recovered as monophyletic, although it is recognized here that color based characters traditionally used for separating the groups were not used in the present work. The subfamily Apiophorinae, with four species sampled, was not recovered as monophyletic, with genus Eumydas grouping to Rhopaliinae. The classification of Mydinae is reviewed here, but due to reasonable uncertainty as to the relationships between some groups, some taxa of the traditional classification were kept, although not recognized as monophyletic
Nunes, Hélio Rubens de Carvalho. "Ponderação Bayesiana de modelos em regressão linear clássica." Universidade de São Paulo, 2005. http://www.teses.usp.br/teses/disponiveis/11/11134/tde-16112005-155133/.
Full textThe objective of this work was divulge to Bayesian Model Averaging (BMA) between the researchers of the agronomy area and discuss its advantages and limitations. With the BMA is possible combine results of difeerent models about determined quantity of interest, with that, the BMA presents as being a metodology alternative of data analysis front the usual models selection approaches, for example the Coefficient of Multiple Determination (R2), Coefficient of Multiple Determination Adjusted (R2), Mallows (Cp Statistics) and Prediction Error Sum Squares (PRESS). Several works recently were carried out with the objective of compare the performance of the BMA regarding the approaches of models selection, however, there is still many situations for will be exploited to that can arrive to a general conclusion about this metodology. In this work, the BMA was applied to data originating from an agronomy experiment. It follow, the predictive performance of the BMA was compared with the performance of the approaches of selection above cited by means of a study of simulation varying the degree of multicollinearity, measured by the number of condition of the matrix standardized X'X and the number of observations in the sample. In each one of those situations, were utilized 1000 samples generated from the descriptive information of agronomy data. The predictive performance of the metodologies in comparison was measured by the Logarithm of the Score Predictive (LEP). The empirical results obtained indicated that the BMA presents similar performance to the usual approaches of selection of models in the situations of multicollinearity exploited.
Fortulan, Viviane Carla. "Meta-Análise: Um Enfoque Bayesiano." Universidade de São Paulo, 1999. http://www.teses.usp.br/teses/disponiveis/55/55134/tde-12032018-110425/.
Full textThis work is concerned with the use of meta-analysis from a Bayesian point of view. Meta-analysis is a quantitative method combining independent results to obtain global conclusions. Frequently different results are combined unapropriately leading to an unreliable inferential analysis. Bayesian methods for meta-analysis are preferred due to the small number of experiments which are caracteristc of this technique. Sets of data taken from the literature, were used to illustrate the use of the technique, and a simulation set. A Bayesian analysis is made possible through the use of Monte Cano simulation methods via Marcos\' Chains.
Forni, Selma. "Análise da curva de crescimento de bovinos da raça Nelore utilizando funções não-lineares em análises Bayesianas / Selma Forni. -." Jaboticabal : [s.n.], 2007. http://hdl.handle.net/11449/102797.
Full textAbstract: The objective of this work was to estimate the joint posterior distribution of Nelore growth curve parameters, their (co)variance components and the environmental and additive genetic components affecting them. The Brody, Von Bertalanffy, Gompertz and Logistic functions were applied in the first stage of a hierarchical Bayesian model. The environmental and genetic effects were described by an animal model in the second stage. Different approaches for describing the adjustment error variance along the growth curve were evaluated: constancy throughout the trajectory, linear increasing until three years of age and exponential increasing. Random samples of the marginal distributions were drawn using Metropolis-Hastings and Gibbs sampling algorithms. Even thought the curve parameters were estimated for animals with records just from the beginning of the growth process, the adult weights were accurately predicted. A high additive genetic variance for mature weight was observed. The parameter a of Brody, Von Bertalanffy and Gompertz models could be used as a selection criterion to control adult weight increases. The effect of maternal environment on growth was carried through to maturity and it should be considered while evaluating all weights. The adjustment error variances at the beginning of growth curve were not adequately described by the linear and exponential models. Selection to change the growth curve slope without modifying adult weight would be inefficient, since their genetic correlation is high.
Orientadora: Lúcia Galvão de Albuquerque
Coorientador: Henrique Nunes de Oliveira
Banca: Joanir Pereira Eler
Banca: Paulo Sávio Lopes
Banca: Humberto Tonhati
Banca: Maurício Mello de Alencar
Doutor
Paraiba, Carolina Costa Mota. "Modelos não lineares truncados mistos para locação e escala." Universidade Federal de São Carlos, 2015. https://repositorio.ufscar.br/handle/ufscar/4497.
Full textFinanciadora de Estudos e Projetos
We present a class of nonlinear truncated mixed-effects models where the truncation nature of the data is incorporated into the statistical model by assuming that the variable of interest, namely the truncated variable, follows a truncated distribution which, in turn, corresponds to a conditional distribution obtained by restricting the support of a given probability distribution function. The family of nonlinear truncated mixed-effects models for location and scale is constructed based on the perspective of nonlinear generalized mixed-effects models and by assuming that the distribution of response variable belongs to a truncated class of distributions indexed by a location and a scale parameter. The location parameter of the response variable is assumed to be associated with a continuous nonlinear function of covariates and unknown parameters and with unobserved random effects, and the scale parameter of the responses is assumed to be characterized by a continuous function of the covariates and unknown parameters. The proposed truncated nonlinear mixed-effects models are constructed assuming both random truncation limits; however, truncated nonlinear mixed-effects models with fixed known limits are readily obtained as particular cases of these models. For models constructed under the assumption of random truncation limits, the likelihood function of the observed data shall be a function both of the parameters of the truncated distribution of the truncated variable and of the parameters of the distribution of the truncation variables. For the particular case of fixed known truncation limits, the likelihood function of the observed data is a function only of the parameters of the truncated distribution assumed for the variable of interest. The likelihood equation resulting from the proposed truncated nonlinear regression models do not have analytical solutions and thus, under the frequentist inferential perspective, the model parameters are estimated by direct maximization of the log-likelihood using an iterative procedure. We also consider diagnostic analysis to check for model misspecification, outliers and influential observations using standardized residuals, and global and local influence metrics. Under the Bayesian perspective of statistical inference, parameter estimates are computed based on draws from the posterior distribution of parameters obtained using an Markov Chain Monte Carlo procedure. Posterior predictive checks, Bayesian standardized residuals and a Bayesian influence measures are considered to check for model adequacy, outliers and influential observations. As Bayesian model selection criteria, we consider the sum of log -CPO and a Bayesian model selection procedure using a Bayesian mixture model framework. To illustrate the proposed methodology, we analyze soil-water retention, which are used to construct soil-water characteristic curves and which are subject to truncation since soil-water content (the proportion of water in soil samples) is limited by the residual soil-water content and the saturated soil-water content.
Neste trabalho, apresentamos uma classe de modelos não lineares truncados mistos onde a característica de truncamento dos dados é incorporada ao modelo estatístico assumindo-se que a variável de interesse, isto é, a variável truncada, possui uma função de distribuição truncada que, por sua vez, corresponde a uma função de distribuição condicional obtida ao se restringir o suporte de alguma função de distribuição de probabilidade. A família de modelos não lineares truncados mistos para locação e escala é construída sob a perspectiva de modelos não lineares generalizados mistos e considerando uma classe de distribuições indexadas por parâmetros de locação e escala. Assumimos que o parâmetro de locação da variável resposta é associado a uma função não linear contínua de um conjunto de covariáveis e parâmetros desconhecidos e a efeitos aleatórios não observáveis, e que o parâmetro de escala das respostas pode ser caracterizado por uma função contínua das covariáveis e de parâmetros desconhecidos. Os modelos não lineares truncados mistos para locação e escala, aqui apresentados, são construídos supondo limites de truncamento aleatórios, porém, modelos não lineares truncados mistos com limites fixos e conhecidos são prontamente obtidos como casos particulares desses modelos. Nos modelos construídos sob a suposição de limites de truncamentos aleatórios, a função de verossimilhança é escrita em função dos parâmetros da distribuição da variável resposta truncada e dos parâmetros das distribuições das variáveis de truncamento. Para o caso particular de limites fixos e conhecidos, a função de verossimilhança será apenas uma função dos parâmetros da distribuição truncada assumida para a variável resposta de interesse. As equações de verossimilhança dos modelos, aqui propostos, não possuem soluções analíticas e, sob a perspectiva frequentista de inferência estatística, os parâmetros do modelo são estimados pela maximização direta da função de log-verossimilhança via um procedimento iterativo. Consideramos, também, uma análise de diagnóstico para verificar a adequação do modelo, observações discrepantes e/ou influentes, usando resíduos padronizados e medidas de influência global e influência local. Sob a perspectiva Bayesiana de inferência estatística, as estimativas dos parâmetros dos modelos propostos são definidas como as médias a posteriori de amostras obtidas via um algoritmo do tipo cadeia de Markov Monte Carlo das distribuições a posteriori dos parâmetros. Para a análise de diagnóstico Bayesiano do modelo, consideramos métricas de avaliação preditiva a posteriori, resíduos Bayesianos padronizados e a calibração de casos para diagnóstico de influência. Como critérios Bayesianos de seleção de modelos, consideramos a soma de log -CPO e um critério de seleção de modelos baseada na abordagem Bayesiana de mistura de modelos. Para ilustrar a metodologia proposta, analisamos dados de retenção de água em solo, que são usados para construir curvas de retenção de água em solo e que estão sujeitos a truncamento pois as medições de umidade de água (a proporção de água presente em amostras de solos) são limitadas pela umidade residual e pela umidade saturada do solo amostrado.
Forni, Selma [UNESP]. "Análise da curva de crescimento de bovinos da raça Nelore utilizando funções não-lineares em análises Bayesianas: Selma Forni. -." Universidade Estadual Paulista (UNESP), 2007. http://hdl.handle.net/11449/102797.
Full textConselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)
O objetivo do presente trabalho foi estimar conjuntamente os parâmetros das curvas de crescimento de animais da raça Nelore, seus componentes de (co)variâncias e os efeitos genéticos e ambientais que atuaram sobre eles. As funções de Brody, Von Bertalanffy, Gompertz e Logística foram empregadas no primeiro estágio de um modelo hierárquico Bayesiano. Os efeitos genéticos e ambientais foram considerados em um modelo animal no segundo estágio de hierarquia. Diferentes abordagens para a variância do erro de ajuste foram avaliadas: constância ao longo da trajetória, aumento linear até os três anos de idade e aumento exponencial. Amostras aleatórias das distribuições marginais foram obtidas aplicando-se os algoritmos de Metropolis-Hastings e amostragem de Gibbs. A presença de animais que não atingiram a maturidade no conjunto de dados não prejudicou a predição dos pesos adultos. Grande parte da variância fenotípica observada neste peso foi devida a efeitos genéticos aditivos. O parâmetro a das curvas de Brody, Von Bertalanffy e Gompertz poderia ser utilizado como critério de seleção para controlar o aumento de peso adulto. O ambiente materno influenciou não somente o crescimento inicial dos animais mas também os pesos maduros e deve ser considerado na avaliação de todas as etapas do crescimento. Os modelos linear e exponencial empregados para a variância do erro de ajuste não representaram de forma adequada este parâmetro no início da curva. A seleção para alterar a pendente da curva de crescimento mantendo o peso adulto constante seria ineficiente, uma vez que, é alta e positiva a correlação genética entre o peso assintótico e a taxa de maturação.
The objective of this work was to estimate the joint posterior distribution of Nelore growth curve parameters, their (co)variance components and the environmental and additive genetic components affecting them. The Brody, Von Bertalanffy, Gompertz and Logistic functions were applied in the first stage of a hierarchical Bayesian model. The environmental and genetic effects were described by an animal model in the second stage. Different approaches for describing the adjustment error variance along the growth curve were evaluated: constancy throughout the trajectory, linear increasing until three years of age and exponential increasing. Random samples of the marginal distributions were drawn using Metropolis-Hastings and Gibbs sampling algorithms. Even thought the curve parameters were estimated for animals with records just from the beginning of the growth process, the adult weights were accurately predicted. A high additive genetic variance for mature weight was observed. The parameter a of Brody, Von Bertalanffy and Gompertz models could be used as a selection criterion to control adult weight increases. The effect of maternal environment on growth was carried through to maturity and it should be considered while evaluating all weights. The adjustment error variances at the beginning of growth curve were not adequately described by the linear and exponential models. Selection to change the growth curve slope without modifying adult weight would be inefficient, since their genetic correlation is high.
CAMPOS, R. G. "Cosmologia Observacional Usando Análise Bayesiana." Universidade Federal do Espírito Santo, 2008. http://repositorio.ufes.br/handle/10/4763.
Full textA cosmologia observacional é baseada em um tripé : dados observacionais, análise estatística e modelos cosmológicos teóricos. Atualmente existe uma grande quantidade de dados cosmológicos observacionais distintos, em nosso trabalho lidamos com : Supernovas do tipo Ia (SNeIa) e fração de gás e massa a partir de raios-x de aglomerados de galáxias (fgas). Como diferencial, utilizamos a elegante e bem fundamentada estatística Bayesiana, em uma análise completa em várias dimensões, com estimativas de parâmetros cosmológicos independentes e dependentes (idade do Universo, parâmetro de desaceleração, etc). A interface entre cosmologia e estatística é feita com a ferramenta computacional por nós desenvolvida, BayEsian Tools for Observational Cosmology (BETOC), que aplica-se facilmente para qualquer modelo cosmológico teórico. Atualmente BETOC possui três variantes, BayEsian Tools for Observational Cosmology using SNeIa-Gold (BETOCS), BayEsian Tools for Observational Cosmology using X-ray of galaxy clusters (BETOCX) e um híbrido de ambos (BETOCSX). Enfim, esse trabalho visa mostrar como a inferência Bayesiana possibilita testes profundos e não tendenciosos nos modelos cosmológicos teóricos, tenham eles poucos parâmetros ou muitos, sejam eles comportados ou não.
Cruz, Ângela Marlene Pires da. "Análise bayesiana de séries temporais." Master's thesis, Universidade de Aveiro, 2008. http://hdl.handle.net/10773/9402.
Full textO presente trabalho propõe-se analisar a metodologia bayesiana para alguns modelos de séries temporais nomeadamente os autoregressivos, os de médias móveis, os inteiros autoregressivos e os inteiros de médias móveis. É feita uma apresentação dos fundamentos Bayesianos, sua aplicação aos modelos referidos, uma exemplificação para cada modelo utilizando um software adequado e uma aplicação da metodologia a dados de grupos de manchas solares de Palehua.
This study proposes to examine the Bayesian approach for some models the of time series including the autoregressive, the moving average, the integer autoregressive and the integer of moving averages. Consists in a presentation of the bayesian methodology, its application to the refered models, an exemplification for each model using an appropriate software and an application of the methodology to the sunspots of Palehua data.
Machado, Fabio Henrique Santana. "Análise de desempenho em redes bayesianas com largura de árvore limitada." Universidade de São Paulo, 2016. http://www.teses.usp.br/teses/disponiveis/3/3141/tde-17042017-093535/.
Full textThis work provides an empirical evaluation of the performance of Bayesian Networks when treewidth is bounded. The performance of the network is viewed as its generalizability and also as the accuracy of inference in decision making problems. Preliminary results suggest that adding constraints to treewidth decreases the model performance on unseen data and makes the corresponding optimization problem more difficult.
Barros, Emilio Augusto Coelho. "Modelagem em análise de sobrevivência para dados médicos bivariados utilizando funções cópulas e fração de cura." Universidade de São Paulo, 2014. http://www.teses.usp.br/teses/disponiveis/17/17139/tde-23092014-120646/.
Full textMixture and non-mixture lifetime models are applied to analyze survival data when some individuals may never experience the event of interest. Dierent statistical models are proposed to analyze survival data in the presence of cure fraction. In this thesis, we propose the use of new models. From the univariate case, we consider that the lifetime data have a three-parameter Burr XII distribution, which includes the popular Weibull mixture model as a special case. We consider a general survival model where the scale and shape parameters of the Burr XII distribution depends on covariates. Also considering the univariate case the two-parameters exponentiated exponential distribution is used. The two-parameter exponentiated exponential or the generalized exponential distribution is a particular member of the exponentiated Weibull distribution introduced by Mudholkar and Srivastava (1993). We also consider in this case a general survival model where the scale, shape and cured fraction parameters of the exponentiated exponential distribution depends on covariates. We also introduce the univariate Weibull distributions in presence of cure fraction, censored data and covariates. Two models are explored in this case: the mixture model and non-mixture model. When we have two lifetimes associated with each unit (bivariate data), we can use some bivariate distributions: as special case the Block and Basu bivariate lifetime distribution. We also presents estimates for the parameters included in Block and Basu bivariate lifetime distribution in presence of covariates and cure fraction, applied to analyze survival data when some individuals may never experience the event of interest and two lifetimes are associated with each unit. We also consider in bivariate case the bivariate Weibull distributions derived from copula functions in presence of cure fraction, censored data and covariates. Two copula functions are explored in this paper: the Farlie-Gumbel-Morgenstern copula (FGM) and the Gumbel copula. Classical and Bayesian procedures are used to get point and condence intervals of the unknown parameters. Illustrations of the proposed methodologies are given considering medicals data sets.
Barros, Emilio Augusto Coelho. "Análise estatística para dados de contagem longitudinais na presença de covariáveis: aplicações na área médica." Universidade de São Paulo, 2009. http://www.teses.usp.br/teses/disponiveis/17/17139/tde-28102009-161433/.
Full textCOELHO-BARROS, E. A. Analise estatstica para dados de contagem longitudinais na presenca de covariaveis: Aplicac~oes na area medica. Dissertac~ao (mestrado) - Faculdade de Medicina de Ribeir~ao Preto - USP, Ribeir~ao Preto - SP - Brasil, 2009. Longitudinal counting data in the presence of covariates is very common in many applications, especially considering medical data. In this work we present dierent \\frailty\"models to analyze longitudinal Poisson data in the presence of covariates. These models incorporate the extra-Poisson variability and the possible correlation among the repeated counting data for each individual. A hierarchical Bayesian analysis is introduced for each dierent model considering usual MCMC (Markov Chain Monte Carlo) methods. Considering reals biological data set (obtained from CEMEQ, Medical School of Ribeir~ao Preto, University of S~ao Paulo, Brazil), we also discuss some Bayesian discrimination aspects for the choice of the best model. In Section 4 is considering a data set related to an open prospective and randomized study, considering of HIV infected patients, free of treatments, which entered the Infection Diseases Therapy Special Unit (UETDI) of the Clinical Hospital of the Medical School of Ribeir~ao Preto, University of S~ao Paulo (HCFMRP-USP). The therapeutic treatments consisted of the drugs Zidovudine and Lamivudine, associated to Efavirenz and Lopinavir. The data set was related to 66 patients followed from September, 2004 to may, 2006, from which, 43 were included in the study. The patients groups presented similar basal characteristics in terms of sex, age, CD4 counting median and viral load. The main goal of this study was to compare the CD4 cells counting for the two treatments, based on the drugs Efavirenz and Lopinavir, recently adopted as preferencial for the initial treatment of the disease.
Obage, Simone Cristina. "Uma análise bayesiana para dados composicionais." Universidade Federal de São Carlos, 2005. https://repositorio.ufscar.br/handle/ufscar/4505.
Full textUniversidade Federal de Sao Carlos
Compositional data are given by vectors of positive numbers with sum equals to one. These kinds of data are common in many applications, as in geology, biology, economy among many others. In this paper, we introduce a Bayesian analysis for compositional data considering additive log-ratio (ALR) and Box-Cox transformations assuming a mul- tivariate normal distribution for correlated errors. These results generalize some existing Bayesian approaches assuming uncorrelated errors. We also consider the use of expo- nential power distributions for uncorrelated errors considering additive log-ratio (ALR) transformation. We illustrate the proposed methodology considering a real data set.
Dados Composicionais são dados por vetores com elementos positivos cuja soma é um. Exemplos típicos de dados desta natureza são encontrados nas mais diversas áreas; como em geologia, biologia, economia entre outras. Neste trabalho, introduzimos uma análise Bayesiana para dados composicionais considerando as transformações razão log-aditiva e Box-Cox, assumindo a distribuição normal multivariada para erros correlacionados. Estes resultados generalizam uma abordagem bayesiana assumindo erros não correlacionados. Também consideramos o uso da distribuição potência exponencial para erros não correla- cionados, assumindo a transformação razão log-aditiva. Nós ilustramos a metodologia proposta considerando um conjunto de dados reais.
Tito, Edison Americo Huarsaya. "Análise de portfólio: uma perspectiva bayesiana." reponame:Repositório Institucional do FGV, 2016. http://hdl.handle.net/10438/16637.
Full textApproved for entry into archive by GILSON ROCHA MIRANDA (gilson.miranda@fgv.br) on 2016-06-23T14:36:07Z (GMT) No. of bitstreams: 1 EdisonMscFGV(20160619).pdf: 2366030 bytes, checksum: 231be2cde1e7f8e01331fddff3f227a1 (MD5)
Approved for entry into archive by GILSON ROCHA MIRANDA (gilson.miranda@fgv.br) on 2016-06-24T12:51:11Z (GMT) No. of bitstreams: 1 EdisonMscFGV(20160619).pdf: 2366030 bytes, checksum: 231be2cde1e7f8e01331fddff3f227a1 (MD5)
Made available in DSpace on 2016-06-29T12:06:48Z (GMT). No. of bitstreams: 1 EdisonMscFGV(20160619).pdf: 2366030 bytes, checksum: 231be2cde1e7f8e01331fddff3f227a1 (MD5) Previous issue date: 2016-06-03
This work has the objective to address the problem of asset allocation (portfolio analysis) under a Bayesian perspective. For this it was necessary to review all the theoretical analysis of the classical mean-variance model and following identify their deficiencies that compromise its effectiveness in real cases. Interestingly, its biggest deficiency this not related to the model itself, but by its input data in particular the expected return calculated on historical data. To overcome this deficiency the Bayesian approach (Black-Litterman model) treat the expected return as a random variable and after that builds a priori distribution (based on the CAPM model) and a likelihood distribution (based on market investor’s views) to finally apply Bayes theorem resulting in the posterior distribution. The expected value of the return of this posteriori distribution is to replace the estimated expected return calculated on historical data. The results showed that the Bayesian model presents conservative and intuitive results in relation to the classical model of mean-variance.
Este trabalho tem com objetivo abordar o problema de alocação de ativos (análise de portfólio) sob uma ótica Bayesiana. Para isto foi necessário revisar toda a análise teórica do modelo clássico de média-variância e na sequencia identificar suas deficiências que comprometem sua eficácia em casos reais. Curiosamente, sua maior deficiência não esta relacionado com o próprio modelo e sim pelos seus dados de entrada em especial ao retorno esperado calculado com dados históricos. Para superar esta deficiência a abordagem Bayesiana (modelo de Black-Litterman) trata o retorno esperado como uma variável aleatória e na sequência constrói uma distribuição a priori (baseado no modelo de CAPM) e uma distribuição de verossimilhança (baseado na visão de mercado sob a ótica do investidor) para finalmente aplicar o teorema de Bayes tendo como resultado a distribuição a posteriori. O novo valor esperado do retorno, que emerge da distribuição a posteriori, é que substituirá a estimativa anterior do retorno esperado calculado com dados históricos. Os resultados obtidos mostraram que o modelo Bayesiano apresenta resultados conservadores e intuitivos em relação ao modelo clássico de média-variância.
Rocha, Everton Batista da. "Modelos para a análise de dados de contagens longitudinais com superdispersão: estimação INLA." Universidade de São Paulo, 2015. http://www.teses.usp.br/teses/disponiveis/11/11134/tde-05112015-144057/.
Full textDiscrete and longitudinal structures naturally arise in clinical trial data. Such data are usually correlated, particularly when the observations are made within the same experimental unit over time and, thus, statistical analyses must take this situation into account. Besides this typical correlation, overdispersion is another common phenomenon in discrete data, defined as a greater observed variability than that nominated by the statistical model. The causes of overdispersion are usually related to an excess of observed zeros (zero-ination), or an excess of observed positive specific values or even both. Molenberghs, Verbeke e Demétrio (2007) have developed a class of models that encompasses both overdispersion and correlation in count data: Poisson, Poisson-gama, Poisson-normal, Poissonnormal- gama (combined model) models. A Bayesian approach was presented by Rizzato (2011) to fit these models using the Markov Chain Monte Carlo method (MCMC). In this work, a Bayesian framework was adopted as well and, in order to consider the uncertainty related to the model parameters, the Integrated Nested Laplace Approximations (INLA) method was used. Along with the models considered in Rizzato (2011), another four new models were proposed including longitudinal correlation, overdispersion and zero-ination by structural and random zeros, namely: zero-inated Poisson (ZIP), zero-inated negative binomial (ZINB), zero-inated Poisson-normal (ZIP-normal) and the zero-inated negative binomial-normal (ZINB-normal) models. In order to illustrate the developed methodology, the models were fit to a real dataset, in which the response variable was taken to be the number of epileptic events per week in each individual. These individuals were split into two groups, one taking placebo and the other taking an experimental drug, and they observed up to 27 weeks. The model selection criteria were given by different predictive measures based on cross validation. In this setting, the ZIP-normal model was selected instead the usual model in the literature (combined model). The computational routines were implemented in R language and constitute a part of this work.
Lizzi, Elisângela Aparecida da Silva. "Padrões espaço-temporais da incidência da AIDS no município de São Paulo, Brasil." Universidade de São Paulo, 2015. http://www.teses.usp.br/teses/disponiveis/17/17139/tde-06012016-140308/.
Full textThis study aims to investigate the spatiotemporal patterns of reported AIDS cases between the years 2000 to 2010 in São Paulo, SP, according to its 96 Administrative Districts, and their associations with sociodemographic characteristics and social vulnerability of these areas. This is an ecological study, and the tools used to incidence rates estimation and analysis of data is Bayesian regression models that incorporate spatial and temporal effects. These models include random effects with a CAR distribution (conditional autoregressive) bivariate normal that capture the influence of adjacent fields on the number of cases reported in each region, according to sex. Estimates of the model parameters were obtained by stochastic simulation method MCMC (Monte Carlo Markov Chain). The spatiotemporal pattern found in this work shows historical traces of the AIDS epidemic in Brazil and the study stratified by gender in the course of the disease for the years under study reveals behavioral aspects of transmission in populations at risk. Can show that the Administrative Districts with better economic classes had higher incidence of the disease in males, since areas with less favored economic classes had higher incidence rates of AIDS among women, justified by the AIDS pauperization and feminization. The results are useful to support the policy planning and health actions directed to the control of AIDS in areas and populations of diverse risk by gender in São Paulo. It emphasizes the need for diversified operations by Administrative District in access to health services, gender, in order to propose a service strategy to separate men and women for areas and providing different support teams to the area with particular characteristics. The bayesian approach proposed proved satisfactory to indicate areas that need more attention as the need for health services, human development and vulnerability.
Polli, Demerson Andre. "Um estudo de métodos bayesianos para dados de sobrevivência com omissão nas covariáveis." Universidade de São Paulo, 2007. http://www.teses.usp.br/teses/disponiveis/45/45133/tde-05052007-174318/.
Full textThe development of methods dealing with missing data is recent in Statistics and is the target of many researchers. The presence of missing values in the covariates is very common in statistical analysis and, in particular, in clinical, epidemiological and enviromental studies for survival data. This work considers a bayesian approach to analise data with missing covariates for parametric models in the Weibull family and for the Cox semiparametric model. The studied methods are evaluated for the parametric and semiparametric approaches considering a dataset of patients with heart insufficiency. Also, the impact of different omission proportions is assessed.
Roza, Daiane Leite da. "Padrões espaço-temporais da incidência da tuberculose em Ribeirão Preto, SP: uso de um modelo bayesiano auto-regressivo condicional." Universidade de São Paulo, 2011. http://www.teses.usp.br/teses/disponiveis/17/17139/tde-20092011-161413/.
Full textIn this study we used Bayesian space-temporal regression models to estimate the incidence of TB in Ribeirão Preto, SP (years 2006 to 2009) by the coverage area of health units, associating it with the covariates of interest (IPVS, Income and Education predominant those areas). The method is based on MCMC simulations for estimate the posterior distributions of TB incidence in Ribeirão Preto. As a result, we have maps that show a spatial pattern more clearly, with estimates smoother and less random fluctuations. We observed that the areas with the highest incidence rates also have medium and high social vulnerability index. Concerning income, the prevailing salary range of household heads in these regions is between 0 and 3 minimum wages and the prevailing level of education of household heads in these regions is the elementary school. The results of the models in Bayesian analysis show that with increasing social vulnerability significantly increased the incidence of TB in Ribeirao Preto. In areas where vulnerability is high incidence of TB is nearly 15 times the incidence of areas without vulnerability. There was a significant increase in the incidence of tuberculosis in Ribeirão Preto during the years studied, the highest incidence recorded in 2009. The use of maps improved visualization of areas that deserve special attention for TB control, in addition, the association of disease with income, education and social vulnerability that bring benefits to the managers responsible for planning the municipality to plan interventions with special attention these areas, uniting efforts to reduce poverty and social inequality, alternatives to improve income distribution and improve access to basic sanitation among other priorities.
Garcia, Lívia Matos [UNESP]. "Análise Bayesiana para a distribuição Exponencial-Logarítmica." Universidade Estadual Paulista (UNESP), 2013. http://hdl.handle.net/11449/94322.
Full textA Exponencial-Logarítmica (EL(p; )) é uma distribuição para modelos de sobrevivência com taxa de falha decrescente. Esta distribuição pode ser usada para estudar os tempos de vida de organismos, materiais, dispositivos, etc, em ciências biológicas e na engenharia. Neste trabalho foram utilizadas as abordagens Clássica e Bayesiana para inferir sobre os parâmetros do modelo com conjunto de dados completos e censurados. A abordagem Bayesiana requer a seleção de distribuições a priori para os parâmetros do modelo. No caso onde há informação dos dados, foram escolhidas distribuições a priori não-informativas. Por outro lado, quando há poucos dados e/ou são dados censurados, torna-se necessária uma priori informativa obtida a partir das informações de um especialista. Neste trabalho propôs-se uma priori informativa elicitada através de informações de especialistas e derivada da aproximação de Laplace. Portanto, a análise Bayesiana foi realizada considerando os dois tipos de distribuição a priori: informativa e não-informativa. As distribuições a priori não-informativas usadas foram: priori de Jeffreys (Jeffreys (1967)), priori de Referência (Berger and Bernardo (1992)), priori de Máxima Informação dos Dados (Zellner (1977)) e priori derivada da função Cópula (Achcar et al. (2010)). Estas distribuições a priori também foram comparadas com outras distribuições a priori comuns, tais como Beta, Gama e Uniforme. A fim de avaliar o desempenho das distribuições a priori, foi apresentado um estudo comparativo utilizando dados simulados a partir da distribuição EL(p; ) e um conjunto de dados reais introduzido por Lawless (1982). Utilizou-se o algoritmo MCMC para obter uma amostra de valores da posteriori conjunta, a fim de extrair características das distribuições posteriores marginais, tais como médias a posteriori, moda e intervalos de credibilidade
The Exponential-Logarithmic, denoted by EL(p; ), is a lifetime distribution with decreasing failure rate. This distribution can be used to study the lengths of organisms, devices, materials, etc., in the biological and engineering sciences. In this dissertation we use classical and Bayesian approaches to make inferences for the parameters of the model under complete and censored data set. Bayesian approach requires the selection of prior distributions for all parameters of the model. In this case, we will seek to choose a noninformative prior that provides best estimation when there is absence of information or with large data set and uncensored data. On the other hand, when there is few data to use or presence of censored data, an informative prior obtained from the expert’s information is necessary. We propose an informative prior elicited from the expert’s opinion and derived though Laplace’s approximation. Thus, we carry out the Bayesian estimation by considering the two types of prior distributions. Different noninformative prior distributions are used as Jeffreys (Jeffreys (1967)), Reference (Berger and Bernardo (1992)), maximal data information prior (Zellner (1977)) and prior derived from copula function (Achcar et al. (2010)). These priors are also compared with other common priors such as beta, gamma, and uniform distributions. A comparative study to evaluate the performance of the prior distributions through simulated data from the EL(p; ) distribution and a practical data set introduced by Lawless (1982) presented. We also need to appeal to the MCMC algorithm to obtain a sample of values of and from the joint posterior in order to extract characteristics of marginal posterior distributions such as Bayes estimator, mode and credible intervals
Garcia, Lívia Matos. "Análise Bayesiana para a distribuição Exponencial-Logarítmica /." Presidente Prudente, 2013. http://hdl.handle.net/11449/94322.
Full textBanca: Josemar Rodrigues
Banca: Sergio Minouru Oikawa
Resumo: A Exponencial-Logarítmica (EL(p; )) é uma distribuição para modelos de sobrevivência com taxa de falha decrescente. Esta distribuição pode ser usada para estudar os tempos de vida de organismos, materiais, dispositivos, etc, em ciências biológicas e na engenharia. Neste trabalho foram utilizadas as abordagens Clássica e Bayesiana para inferir sobre os parâmetros do modelo com conjunto de dados completos e censurados. A abordagem Bayesiana requer a seleção de distribuições a priori para os parâmetros do modelo. No caso onde há informação dos dados, foram escolhidas distribuições a priori não-informativas. Por outro lado, quando há poucos dados e/ou são dados censurados, torna-se necessária uma priori informativa obtida a partir das informações de um especialista. Neste trabalho propôs-se uma priori informativa elicitada através de informações de especialistas e derivada da aproximação de Laplace. Portanto, a análise Bayesiana foi realizada considerando os dois tipos de distribuição a priori: informativa e não-informativa. As distribuições a priori não-informativas usadas foram: priori de Jeffreys (Jeffreys (1967)), priori de Referência (Berger and Bernardo (1992)), priori de Máxima Informação dos Dados (Zellner (1977)) e priori derivada da função Cópula (Achcar et al. (2010)). Estas distribuições a priori também foram comparadas com outras distribuições a priori comuns, tais como Beta, Gama e Uniforme. A fim de avaliar o desempenho das distribuições a priori, foi apresentado um estudo comparativo utilizando dados simulados a partir da distribuição EL(p; ) e um conjunto de dados reais introduzido por Lawless (1982). Utilizou-se o algoritmo MCMC para obter uma amostra de valores da posteriori conjunta, a fim de extrair características das distribuições posteriores marginais, tais como médias a posteriori, moda e intervalos de credibilidade
Abstract: The Exponential-Logarithmic, denoted by EL(p; ), is a lifetime distribution with decreasing failure rate. This distribution can be used to study the lengths of organisms, devices, materials, etc., in the biological and engineering sciences. In this dissertation we use classical and Bayesian approaches to make inferences for the parameters of the model under complete and censored data set. Bayesian approach requires the selection of prior distributions for all parameters of the model. In this case, we will seek to choose a noninformative prior that provides best estimation when there is absence of information or with large data set and uncensored data. On the other hand, when there is few data to use or presence of censored data, an informative prior obtained from the expert's information is necessary. We propose an informative prior elicited from the expert's opinion and derived though Laplace's approximation. Thus, we carry out the Bayesian estimation by considering the two types of prior distributions. Different noninformative prior distributions are used as Jeffreys (Jeffreys (1967)), Reference (Berger and Bernardo (1992)), maximal data information prior (Zellner (1977)) and prior derived from copula function (Achcar et al. (2010)). These priors are also compared with other common priors such as beta, gamma, and uniform distributions. A comparative study to evaluate the performance of the prior distributions through simulated data from the EL(p; ) distribution and a practical data set introduced by Lawless (1982) presented. We also need to appeal to the MCMC algorithm to obtain a sample of values of and from the joint posterior in order to extract characteristics of marginal posterior distributions such as Bayes estimator, mode and credible intervals
Mestre
Lima, Rodolfo Valentim da Costa. "Análise Bayesiana de dois problemas em Astrofísica Relativística: neutrinos do colapso gravitacional e massas das estrelas de nêutrons." Universidade de São Paulo, 2012. http://www.teses.usp.br/teses/disponiveis/14/14131/tde-26062013-162012/.
Full textThe extraordinary event of supernova has been investigated twenty five years ago. The fascination surrounds such astronomical event is on the real time observation the explosion at light to neutrino Physics. Detectors spread for the world had observed one burst neutrinos that days later it was confirmed as being of SN1987A. Kamiokande, IMB and Baksan had presented the detected events that allowed to the study of models for the explosion and cooling of hypothetical neutron star remain. Until today it does not have a consensus the origin of the progenitor and the nature of the remaining compact object. The work is divided in two parts: study of the neutrinos of SN1987A through Analysis Bayesiana Statistics through a model considered with two temperatures that two evidence bursts of neutrinos. The motivation is in the hypothesis of as burst as resulted of the formation of strange matter in the compact object. The employed methodology was developed for an interesting work of Loredo & Lamb (2002) that it allows shape and to test hypotheses on the models saw Bayesian Information Criterion (BIC). The second part of the work, the same methodology statistics is used in the study of the distribution of masses of the neutron stars using the available database http://stellarcollapse.org/. The database was analyzed only using the value of the object and its shunting line standard. Constructing to a a priori function likelihood and using distributions with hypothesis of bimodal distribution of the masses against a unimodal distribution on all the masses of objects. Test BIC indicates fort favorable trend the existence of the bimodality with values centered in 1.37M for objects of low mass and 1.73M for objects of high mass and week evidence of one third peak around 1.25M.
Greni, Susan Elaine. "Análise multigênica e distribuição especial de espécies do Subgrupo Strodei de Anopheles (Nyssorhynchus) (Diptera: Culicidae)." Universidade de São Paulo, 2016. http://www.teses.usp.br/teses/disponiveis/6/6132/tde-05012017-094517/.
Full textIntroduction Anopheles strodei sensu lato is an understudied subgroup of potential epidemiological importance, having been found naturally infected in Brazil with Plasmodium falciparum, Plasmodium vivax and Plasmodium malariae. An. strodei s.l. is currently composed on 8 species: An. albertoi Unti, An. CP Form, An. rondoni (Neiva & Pinto), An. strodei Root, An. arthuri Unti and three other unnamed species that have been proposed by Bourke et al. (2013): An. arthuri B, An. arthuri C and An. arthuri D. Objectives As delineating species accurately is an essential goal of public health entomology, the objectives of this study were to: 1) Determine the phylogenetic relationships within the Strodei Subgroup and reaffirm or reject the hypothesis of the 3 new species (An. arthuri B, An. arthuri C and An. arthuri D) 2) Address the potential spatial distribution of species of the An. strodei subgroup to provide support for the candidate species in the Strodei Subgroup Methods Bayesian inference, which included DNA sequences of one mitochondrial and three nuclear protein coding genes: CO1, white, CAD and CAT, was used to determine the phylogenetic relationship within the group. To propose a species distribution, collection localities, along with climatic and geographic data were input into MAXENT. Results When analyzing the four molecular markers employed, support was found for allopatry in the Strodei Subgroup. The paraphyletic clade of An. arthuri was supported. Conclusion Potential species distributions of the Strodei Subgroup were addressed for the first time. Fifty-five unique CAT sequences and 46 unique CAD sequences were newly characterized.
Silva, Nélia Maria Marques da. "Análise bayesiana de séries temporais de valores inteiros." Doctoral thesis, Universidade de Aveiro, 2005. http://hdl.handle.net/10773/4571.
Full textModelar senes temporais de valores inteiros não negativos (ou senes temporais de contagem) pareceu-nos um desafio bastante aliciante, não só devido à sua importância, como também ao facto de ser um tema ainda pouco explorado, contrariamente à modelação de séries temporais com suporte nos reais. Vários modelos para processos estacionários com distribuição marginal discreta têm sido propostos. Um desses modelos particularmente usado para séries de contagem é o processo Auto-Regressivo de valores inteiros de ordem p, designado por INAR(p). De uma forma geral, este trabalho tem como principal objectivo desenvolver uma abordagem bayesiana aos problemas da estimação de parâmetros e da predição de observações futuras, do ponto de vista pontual e intervalar, nos modelos INAR. Simultaneamente é feita um estudo comparativo com a abordagem clássica. Consideram-se modelos baseados no processo Auto-Regressivo de valores inteiros de 1 a ordem com distribuição marginal de Poisson, designados abreviadamente por PoINAR(1). Inicialmente o estudo incide sobre o modelo PoINAR(1), depois são consideradas réplicas independentes desse modelo e firü~IrTfente,na6 íiiipõnaó-a exi~fênêll:r deinâependência-entre- obs-ervações; faz,;; se uma generalização do modelo PoINAR(1) para um painel com r unidades e n períodos de tempo. Os modelos considerados são comparados e ilustrados através de estudos de simulação e aplicados a dados reais. Todo o trabalho desenvolvido envolve grandes exigências computacionais recorrendo-se intensivamente a alguns dos mais actuais métodos de simulação, com natural destaque para os métodos de simulação de Monte Carla via cadeias de Markov.
Modelling non-negative integer-valued time series is, at the moment, a challenge, not only because of their applicability but also beca use it is still much of an open problem. In the last decades, severa 1 models for stationary processes with discrete marginal distribution have been proposed in the literature. One of the most promising models for time series of counts is the Integer-valued AutoRegressive, INAR, processo The purpose of this essay is to develop a Bayesian approach to the problems of estimation and prediction of future observations in I NAR processes. A comparative study between Bayesian and classical methodologies is carried out. Several models based on the INAR process are considered, the first of which is the first order INAR model with Poisson marginal distribut;on, denoted PoINAR(1). This process ;s then considered in the context of a panel of r independent replicates. Later, the hypothesis of independence between observed units is dropped, resulting in a model denoted by SUINAR(1). Results are iIIustrated using simulation studies and real data sets. Ali the work presented in this thesis, involves a considerable amount of computation -üsln-g--som-e- of -the -most--current simulation metheds fQGusing, naturally, on the Markov chain Monte Carlo methods.
Janeiro, Vanderly. "Análise Bayesiana de Modelos para Dados Binários Correlacionados." Universidade de São Paulo, 2000. http://www.teses.usp.br/teses/disponiveis/55/55134/tde-02032018-141504/.
Full textIn this dissertation, we develop a Bayesian analysis of regression models for correlated binary data in the presence of covariates, including the case with replicates. We consider probit and logistic regression models for correlated binary data assuming random effects with a mixture of normal distributions, since this model have great flexibility to the fitted for correlated binary data. We also present some considerations for the case with replicates. We assume informative prior distributions for the parameters of the model and we use Gibbs sampling and Metropolis-Hastings algorithms to get Monte Cano estimates for the posterior quantities of interest. We also present some considerations for the selection of models using discrepancy measures between the fitted model and the data (Pearson residuais) and using the predictive densities (Bayes factor) estimated by MCMC (Markov Chain Monte Cano). We present a numerical example to illustrate the proposed methodology.
Mazucheli, Josmar. "ANÁLISE BAYESIANA E DISCRIMINAÇÃO DE MODELOS NÃO LINEARES." Universidade de São Paulo, 1995. http://www.teses.usp.br/teses/disponiveis/55/55134/tde-18042018-103011/.
Full textIt is common in many scientific applications, the existence of different non-linear regression models to be used in the same problem. Therefore, usually the researcher has a question: Which model is preferable? This is a question concemed by many statisticians, and many classical or Bayesian strategies for discrimination have been proposed in the literature. In this work, considering the logistic, Gompertz, Weibull-type, Morgan-Mercer- Flodin and Richards growth non-linear models, we present sorne existing strategies to be used in the discrimination of altemative models. Under the classical approach, the discrimination is based on non-linearity concepts, since the best model among many existing altematives is the one that presents behavior close to linear models. Under the Bayesian approach, considering Jeffreys non informative prior densities end Laplace\'s method for approximation of integrals, and a general discrimination procedure, (see Gelfand and Dey, 1994), we explore in an example some different discrimination strategies: Bayes Factor, Entropy, Pseudo Factor of Bayes and Posterior Bayes Factor.
Marques, Katia Antunes. "Análise bayesiana em modelos TRI de três parâmetros." Universidade de São Paulo, 2008. http://www.teses.usp.br/teses/disponiveis/45/45133/tde-02092008-214645/.
Full textIn this dissertation the bayesian analysis for three parameters IRT (Item Response Theory) models with binaries and ordinals responses, considering the probit model, was discussed. For both cases, binary and ordinal, techniques based on MCCM (Monte Carlo Markov Chain) were used to estimate the items parameters. For binary response model, was considered two data sets from tests with multipla choices items. For these two data sets, a sensibility study of the priori distributions choice was considered, and also, an analyses of a posteriori estimates of the items parameters: discrimination, difficulties and guessing. A third data set is used to ilustrate the ordinal response model. This come from an elementar statistical course, where a test with open items is considered. The responses are classified in the following categories: correct, wrong or partial correct. The WinBugs software was used to estimate the parameters for the binary model and, for the ordinal model was considered the function MCMCordfactanal from R program.
Dias, Junior Avelino Viana. "Análise estatística bayesiana em processos com longa dependência." reponame:Biblioteca Digital de Teses e Dissertações da UFRGS, 2010. http://hdl.handle.net/10183/115499.
Full textThe Bayesian approach in statistical inference has been widely used as an alternative to traditional methods. In this work, we present a Bayesian approach for estimating the parameters of the autoregressive moving average processes of orderp and q, denoted by ARMA(p, g) and of the autoregressive fractionally integrated moving average process, denoted by ARFIMA(p, d, g). For the later model, the Bayesian approach is performed assuming p = g = 0. Whereas AR(p), which is a particular case of the ARMA(p, g) model when g = O, an estimator is proposed via the Bayesian approach. The efficiency of the estimator is verified by Monte Cario simulations and the results are compared with the classical maximum likelihood estimator. In the case of ARFIMA(0, d, 0) process, a theoretical study is performed by the Bayesian approach. For estimating the parameters of that process we consider its infiriite autoregressive representation. Some Bayesian computational algorithms are presented in this work since they play an important role in Bayesian inferences. Some of these algorithms, such as Gibbs sampler and Metropolis-Hastings algorithm, were used in building the estimators for the parameters of ARMA and ARFIMA models.
Saraiva, Sandra Maria Bargão. "Análise discriminante, teoria da decisão e inferência bayesiana." Master's thesis, Universidade de Évora, 2000. http://hdl.handle.net/10174/13430.
Full textChristiansen, Trujillo Andrés Guillermo. "Análisis de influencia bajo inferencia bayesiana en evaluaciones escolares de altas consecuencias." Master's thesis, Pontificia Universidad Católica del Perú, 2018. http://tesis.pucp.edu.pe/repositorio/handle/123456789/12356.
Full textTesis
Nunes, André Francisco Nunes de. "Políticas monetária e fiscal ativas e passivas : uma análise para o Brasil pós-metas de inflação." reponame:Biblioteca Digital de Teses e Dissertações da UFRGS, 2009. http://hdl.handle.net/10183/18811.
Full textThis paper seeks identify whether the way of fiscal and monetary macroeconomic policies in Brazil, to that period after inflation targets, were active way or/and passive way. For that, it’s estimated, for Bayesian methods, a model DSGE with price rigidities and monopolistic competition, in which the primary surplus and the nominal interest rates are the tools economic policy available. The lack of coordination of policies in Brazil, usually, has been identified as the reason for the macroeconomic imbalances. So, many authors pointed out the active fiscal policy, as a factor limiting the efficient performance of monetary policy. However, the analysis that relation within the framework of DSGE models is still limited, especially in applications for the Brazilian economy. The estimates of the model pointed out for a system where policies were active during the 2000/1Q to 2002/4Q both of them, and the later period, 2003/1Q – 2008/4Q, the fiscal policy behaved themselves on passive way and the monetary policy was active way.
Vencio, Ricardo Zorzetto Nicoliello. "Análise estatística na interpretação de imagens: microarranjos de DNA e ressonância magnética funcional." Universidade de São Paulo, 2006. http://www.teses.usp.br/teses/disponiveis/95/95131/tde-16032007-164424/.
Full textThe goal of this work is to present the novel Bioinformatics methods that were developed aiming the statistical analysis of two image-based techniques: DNA microarrays and functional magnetic resonance imaging. The main interest is to approach these experimental techniques in small sample size situations, i.e., when there are relatively few experimental observations of the phenomena of interest, for which the case of single subject/datum analysis is its most extreme. In order to approach these problems we chose to use Bayesian Inference in the context of the Decision Theory under Uncertainty, computationally implemented under the Decision Support Systems framework. Both technologies produce complex data, based on the interpretation of differences between images from the response to a given stimulus and the control situation. The result of this work is the development of two decision support systems, called HTself and Dotslashen, to analyze microarray and functional magnetic resonance imaging data, respectively; and the underling mathematical and computational methods. These systems use the rational knowledge from normative databases implemented in specific mathematical models, overcoming the problem of small sample size. Finally, in this work it is described applications to real problems in order to stress the utility for Molecular Biology and Functional Neuroimaging of the developed decision support systems.
Paixão, Fábio de Araújo Jesus. "Estimação bayesiana via cópulas para dados com censura intervalar bivariados." reponame:Repositório Institucional da UnB, 2015. http://dx.doi.org/10.26512/2015.06.D.18873.
Full textSubmitted by Tania Milca Carvalho Malheiros (tania@bce.unb.br) on 2015-10-27T13:43:09Z No. of bitstreams: 1 2015_FábiodeAraújoJesusPaixão.pdf: 592208 bytes, checksum: 49115036f169da5cc3b62689cee5fcff (MD5)
Approved for entry into archive by Patrícia Nunes da Silva(patricia@bce.unb.br) on 2015-12-04T14:53:10Z (GMT) No. of bitstreams: 1 2015_FábiodeAraújoJesusPaixão.pdf: 592208 bytes, checksum: 49115036f169da5cc3b62689cee5fcff (MD5)
Made available in DSpace on 2015-12-04T14:53:10Z (GMT). No. of bitstreams: 1 2015_FábiodeAraújoJesusPaixão.pdf: 592208 bytes, checksum: 49115036f169da5cc3b62689cee5fcff (MD5)
Neste trabalho apresentamos uma metodologia bayesiana em dois est agios para estimar a fun ção de distribui ção conjunta entre tempos de sobrevivência bivariados, com censura intervalar. A estrutura de dependência das vari áveis foi representada por um modelo c ópula, em particular usando uma função da familia das cópulas arquimedianas. As distribui ções marginais foram modeladas assumindo tempos de falha simulados com distribui ção Weibull e ajustados seguindo o contexto bayesiano, utilizando o m étodo de simula ção MCMC (Monte Carlo via Cadeias de Markov) Metropolis-Hastings. Para a estima ção da fun ção de distribui ção conjunta foram simuladas amostras da distribui ção a posteriori conjunta, obtida via metodologia bayesiana com uso de fun ção c ópula. Os resultados experimentais demonstram que o m étodo proposto fornece estimativas satisfat órias quando os modelos marginais e de c ópulas são supostos corretamente. Todo o c ódigo foi implementado utilizando o software estatístico livre R.
We present a Bayesian methodology in two stages to estimate the joint distribution function of bivariate interval censored survival data. The dependence structure of the variables was represented by a copula model, in particular by using a Archimedean copula. The marginal distributions were modeled assuming simulated failure times with Weibull distribution and adjusted using the Bayesian framework, using the Metropolis-Hastings MCMC simulation method (Markovchain Monte Carlo). To get the bivariate distribution function we need sample from the posterior distribution, gotten by Bayesian methodology using copulas. The experimental results demonstrate that the proposed method provides good estimates when the marginal and copulas models are supposed correctly. Al lthe code was implemented using the free statistical software R.
Biz, Guilherme. "Análise Bayesiana de ensaios fatoriais 2k usando os princípios dos efeitos esparsos, da hierarquia e da hereditariedade." Universidade de São Paulo, 2010. http://www.teses.usp.br/teses/disponiveis/11/11134/tde-24022010-092341/.
Full textIn experimental planning for adjustment of polynomials models involving k main factors and their interactions, it is frequent to adopt the 2k, 3k designs or its fractions. Furthermore, it is not unusual, when analysing the results of such experiments, to consider the heredity principle. In other words, once detected a signicant interaction between factors, the factors that appear in this interaction and respective interactions should also be present in the model. In this work, this principle is incorporated directly in the prior, following the ideas proposed by Chipman, Hamada and Wu (1997), but changing some of the hyperparameters. What improves considerably the original methodology. Finally the methodology is illustrated by the analysis of the results of an experiment for the elaboration of pea starch biolms.
Agurto, Mejía Hugo Miguel. "Inferencia bayesiana en un modelo de regresión cuantílica semiparamétrico." Master's thesis, Pontificia Universidad Católica del Perú, 2013. http://tesis.pucp.edu.pe/repositorio/handle/123456789/6174.
Full textTesis
Melani, Arthur Henrique de Andrade. "Desenvolvimento de um método para diagnose de falhas na operação de navios transportadores de gás natural liquefeito através de redes bayesianas." Universidade de São Paulo, 2015. http://www.teses.usp.br/teses/disponiveis/3/3151/tde-15062016-154821/.
Full textLiquefied Natural Gas (LNG) has gradually become an important option for the diversification of the Brazilian energy matrix. LNG carriers are responsible for LNG transportation from the liquefaction plant to the regaseification plant. Given the importance, as well as the risk, of transportation and loading/unloading operations of LNG carriers, not only a good maintenance plan is needed, but also a failure detection system that localizes the origin of a failure that may occur during these processes. This research presents a fault diagnosis method for the loading and unloading operations of LNG carriers through the use of Bayesian networks together with reliability analysis techniques, such as Failure Modes and Effects Analysis (FMEA ) and Fault Tree Analysis (FTA). The proposed method indicates, by reading sensors present in the loading and unloading system, which components are most likely faulty. The method provides a well-structured approach for the development of Bayesian networks used in the diagnosis of system failures.
Niiyama, Clóvis Augusto [UNESP]. "Análise clássica e bayesiana do modelo Weibull modificado generalizado." Universidade Estadual Paulista (UNESP), 2013. http://hdl.handle.net/11449/94326.
Full textNa literatura existem varias distribuições de probabilidade utilizadas em confiabilida e analise de sobrevivência. Entre as famílias de distribuições utilizadas para este fim é a mais popular e a distribuição de Weibull cuja fução de risco apresenta formas: constante, crescente e decrescente. No entanto, quando a função de risco e do tipo unimodal ou em forma de banheira, a Weibull distribui c~ao n~ao e apropriada. Assim, nos ultimos anos, t em sido propostas novas distribui c~oes que acomodam as v arias formas que a fun c~ao de risco pode tomar e consequentemente, para se ajustar a um maior n umero de problemas pr aticos. Carrasco et al. (2008) prop os uma nova distribui c~ao chamada Weibull Modi cada Generalizada, denotada por WMG, sua fun c~ao de risco pode assumir muitas formas, tais como constante, crescente, decrescente, unimodal e banheira. A distribui c~ao Weibull Modi cada Generalizada proposta por Carrasco, Ortega e Cordeiro (2008) foi amplamente estudada no contexto de infer encia cl assica, por em n~ao existem ainda trabalhos desenvolvidos na literatura sob o enfoque Bayesiano. O objetivo deste trabalho foi realizar uma compara c~ao entre os m etodos de estima c~ao cl assico e Bayesiano para a distribui c~ao Weibull Modi cada Generalizada. Tal distribui c~ao ainda tem como sub-modelos as distribui c~oes Exponencial, Exponencial Generalizada, Weibull, Weibull Modi cada, Weibull Exponenciada e valor extremo. Foram realizados estudos sobre as propriedades da distribui c~aoWeibull Modi cada Generalizada e simula c~oes para comparar o desempenho dos estimadores de m axima verossimilhan ca e Bayesiano. Uma abordagem Cl assica e Bayesiana para a esta distribui c~ao foi proposta e exempli cada, modelando conjunto de dados de sobreviv encia e de con abilidade
In the literature there are various probability distributions to model lifetimes of equipment or individual problems in survival analysis. Among the families of distributions used for this purpose, the most popular is the Weibull distribution whose hazard function presents constant, increasing and decreasing forms. However, when the hazard function is the type unimodal or bathtub shaped, the Weibull distribution is not appropriated. Thus, in recent years, there have been proposed new distributions that t the various forms that the hazard function can take and consequently to t a greater number of practical problems. Carrasco et al. (2008) has proposed a new distribution called Generalized Modi ed Weibull, denoted by GMW, whose hazard function can take many forms such as constant, increasing, decreasing, unimodal and bathtub. The Generalized Modi ed Weibull distribution proposed by Carrasco, Ortega e Cordeiro (2008) was most studied in the context of classical inference. However, no studies were found under the Bayesian approach. The aim of this work was to do a comparison of estimation methods for classical and Bayesian Generalized Modi ed Weibull distribution. The Generelized Modi ed Weibull distribuition has a function of risk that can be increasing, decreasing, unimodal and bathtub shaped and has as sub-models Exponential distributions, Exponentiated Exponential, Weibull, Modi ed Weibull, Exponentiated Weibull and extreme value. It was performed the properties of Generalized Modi ed Weibull distribution and a simulation study to compare the performance of maximum likelihood estimator and Bayesian estimator. Classical and Bayesian approach to this distribution was proposed and exempli ed, modeling data sets of survival and reliability
Shimizu, Taciana Kisaki Oliveira [UNESP]. "Análise Bayesiana de dados composicionais na presença de covariáveis." Universidade Estadual Paulista (UNESP), 2014. http://hdl.handle.net/11449/108634.
Full textDados composicionais consistem em vetores conhecidos como composições cujos componentes são positivos e definidos no intervalo (0,1) representando proporções ou frações de um “todo”. A soma desses componentes deve ser igual a um. Os dados composicionais estão presentes em diferentes áreas, como na geologia, ecologia, economia, medicina entre muitas outras. Desta forma há um grande interesse em novas abordagens de modelar dados composicionais. Neste estudo, introduzimos as transformações logaritmo da razão (alr) e Box-Cox em modelos usados para dados composicionais, assumindo erros normais não correlacionados. O objetivo principal deste trabalho é aplicar métodos Bayesianos para estes modelos utilizando os métodos padrões de Monte Carlo via Cadeias de Markov (MCMC) para simular amostras da posteriori conjunta de interesse. Nós aplicamos a metodologia proposta em dois conjuntos de dados, sendo que um deles é sobre um experimento de medidas repetidas na qual introduzimos uma variável de efeito aleatório para capturar a dependência para os dados longitudinais e, além disso, a introdução de dois efeitos aleatórios extras no modelo. Estes resultados de modelagem podem ser de grande interesse em trabalhos aplicados que lidam com conjuntos de dados composicionais.
Compositional data consist of known compositions vectors whose components are positive and defined in the interval (0,1) representing proportions or fractions of a “whole”. The sum of these components must be equal to one. Compositional data is present in different areas, as in ecology, economy, medicine among many others. In this way, there is a great interest in new modeling approaches for compositional data. In this study we introduced additive log-ratio (alr) and Box-Cox transformations models used for compositional data, under uncorrelated normal errors. The main objective of this project is to apply Bayesian methods to these models using standard Markov Chain Monte Carlo (MCMC) methods to simulate samples of the joint posterior of interest. We apply the proposed methodology in two data sets, whereas one of them is about an experiment of repeated measures where we introduced a random effect variable to capture the dependence for the longitudinal data and also the introduction of two extra random effects in the model. These modeling results could be of great interest in the applied work dealing with compositional data sets.
Shimizu, Taciana Kisaki Oliveira. "Análise Bayesiana de dados composicionais na presença de covariáveis /." Presidente Prudente, 2014. http://hdl.handle.net/11449/108634.
Full textCoorientador: Mário Hissamitsu Tarumoto
Banca: Carlos Aparecido dos Santos
Banca: Aparecida Doniseti Pires de Souza
Resumo: Dados composicionais consistem em vetores conhecidos como composições cujos componentes são positivos e definidos no intervalo (0,1) representando proporções ou frações de um "todo". A soma desses componentes deve ser igual a um. Os dados composicionais estão presentes em diferentes áreas, como na geologia, ecologia, economia, medicina entre muitas outras. Desta forma há um grande interesse em novas abordagens de modelar dados composicionais. Neste estudo, introduzimos as transformações logaritmo da razão (alr) e Box-Cox em modelos usados para dados composicionais, assumindo erros normais não correlacionados. O objetivo principal deste trabalho é aplicar métodos Bayesianos para estes modelos utilizando os métodos padrões de Monte Carlo via Cadeias de Markov (MCMC) para simular amostras da posteriori conjunta de interesse. Nós aplicamos a metodologia proposta em dois conjuntos de dados, sendo que um deles é sobre um experimento de medidas repetidas na qual introduzimos uma variável de efeito aleatório para capturar a dependência para os dados longitudinais e, além disso, a introdução de dois efeitos aleatórios extras no modelo. Estes resultados de modelagem podem ser de grande interesse em trabalhos aplicados que lidam com conjuntos de dados composicionais.
Abstract: Compositional data consist of known compositions vectors whose components are positive and defined in the interval (0,1) representing proportions or fractions of a "whole". The sum of these components must be equal to one. Compositional data is present in different areas, as in ecology, economy, medicine among many others. In this way, there is a great interest in new modeling approaches for compositional data. In this study we introduced additive log-ratio (alr) and Box-Cox transformations models used for compositional data, under uncorrelated normal errors. The main objective of this project is to apply Bayesian methods to these models using standard Markov Chain Monte Carlo (MCMC) methods to simulate samples of the joint posterior of interest. We apply the proposed methodology in two data sets, whereas one of them is about an experiment of repeated measures where we introduced a random effect variable to capture the dependence for the longitudinal data and also the introduction of two extra random effects in the model. These modeling results could be of great interest in the applied work dealing with compositional data sets.
Mestre
Niiyama, Clóvis Augusto. "Análise clássica e bayesiana do modelo Weibull modificado generalizado /." Presidente Prudente, 2013. http://hdl.handle.net/11449/94326.
Full textBanca: Mario Hissamitsu Tarumoto
Banca: Roseli Aparecida Leandro
Resumo: Na literatura existem varias distribuições de probabilidade utilizadas em confiabilida e analise de sobrevivência. Entre as famílias de distribuições utilizadas para este fim é a mais popular e a distribuição de Weibull cuja fução de risco apresenta formas: constante, crescente e decrescente. No entanto, quando a função de risco e do tipo unimodal ou em forma de banheira, a Weibull distribui c~ao n~ao e apropriada. Assim, nos ultimos anos, t^em sido propostas novas distribui c~oes que acomodam as v arias formas que a fun c~ao de risco pode tomar e consequentemente, para se ajustar a um maior n umero de problemas pr aticos. Carrasco et al. (2008) prop^os uma nova distribui c~ao chamada Weibull Modi cada Generalizada, denotada por WMG, sua fun c~ao de risco pode assumir muitas formas, tais como constante, crescente, decrescente, unimodal e banheira. A distribui c~ao Weibull Modi cada Generalizada proposta por Carrasco, Ortega e Cordeiro (2008) foi amplamente estudada no contexto de infer^encia cl assica, por em n~ao existem ainda trabalhos desenvolvidos na literatura sob o enfoque Bayesiano. O objetivo deste trabalho foi realizar uma compara c~ao entre os m etodos de estima c~ao cl assico e Bayesiano para a distribui c~ao Weibull Modi cada Generalizada. Tal distribui c~ao ainda tem como sub-modelos as distribui c~oes Exponencial, Exponencial Generalizada, Weibull, Weibull Modi cada, Weibull Exponenciada e valor extremo. Foram realizados estudos sobre as propriedades da distribui c~aoWeibull Modi cada Generalizada e simula c~oes para comparar o desempenho dos estimadores de m axima verossimilhan ca e Bayesiano. Uma abordagem Cl assica e Bayesiana para a esta distribui c~ao foi proposta e exempli cada, modelando conjunto de dados de sobreviv^encia e de con abilidade
Abstract: In the literature there are various probability distributions to model lifetimes of equipment or individual problems in survival analysis. Among the families of distributions used for this purpose, the most popular is the Weibull distribution whose hazard function presents constant, increasing and decreasing forms. However, when the hazard function is the type unimodal or bathtub shaped, the Weibull distribution is not appropriated. Thus, in recent years, there have been proposed new distributions that t the various forms that the hazard function can take and consequently to t a greater number of practical problems. Carrasco et al. (2008) has proposed a new distribution called Generalized Modi ed Weibull, denoted by GMW, whose hazard function can take many forms such as constant, increasing, decreasing, unimodal and bathtub. The Generalized Modi ed Weibull distribution proposed by Carrasco, Ortega e Cordeiro (2008) was most studied in the context of classical inference. However, no studies were found under the Bayesian approach. The aim of this work was to do a comparison of estimation methods for classical and Bayesian Generalized Modi ed Weibull distribution. The Generelized Modi ed Weibull distribuition has a function of risk that can be increasing, decreasing, unimodal and bathtub shaped and has as sub-models Exponential distributions, Exponentiated Exponential, Weibull, Modi ed Weibull, Exponentiated Weibull and extreme value. It was performed the properties of Generalized Modi ed Weibull distribution and a simulation study to compare the performance of maximum likelihood estimator and Bayesian estimator. Classical and Bayesian approach to this distribution was proposed and exempli ed, modeling data sets of survival and reliability
Mestre