Submitted by Joaquim Heck (jheck@embrioconsult.com.br) on 2012-09-26T17:54:24Z No. of bitstreams: 1 PDF_MPFE_2010_Dissertacao_Joaquim_Heck_2012_08_31.pdf: 1254577 bytes, checksum: 286251ccfbb405513a08d5b0d097a172 (MD5) / Approved for entry into archive by Suzinei Teles Garcia Garcia (suzinei.garcia@fgv.br) on 2012-09-26T18:01:33Z (GMT) No. of bitstreams: 1 PDF_MPFE_2010_Dissertacao_Joaquim_Heck_2012_08_31.pdf: 1254577 bytes, checksum: 286251ccfbb405513a08d5b0d097a172 (MD5) / Made available in DSpace on 2012-09-26T18:09:46Z (GMT). No. of bitstreams: 1 PDF_MPFE_2010_Dissertacao_Joaquim_Heck_2012_08_31.pdf: 1254577 bytes, checksum: 286251ccfbb405513a08d5b0d097a172 (MD5) Previous issue date: 2012-08-31 / This thesis discusses aspects of the demand for healthcare in the Brazilian private health sector. We use econometric analysis of count data models to establish which monetary and non-monetary parameters may influence the demand of healthcare. Finally, we verify if there is any informational asymmetry effect such as moral hazard in the determination of the demand for a case-study involving medical speciality visits. / Este ensaio apresenta um estudo sobre a demanda por serviço de saúde no mercado de saúde suplementar utilizando, através de uma análise econométrica, modelos de regressão de dados de contagem para verificar os fatores monetários e não monetários que podem influenciar a quantidade demandada por este serviço, e determinar se há risco moral na determinação desta demanda, no caso de um modelo de visitas médicas de especialidade.

In this thesis, we consider several statistical techniques applicable to claim frequency models of an MTPL portfolio with a focus on overdispersion. The practical part of the work is focused on the application and comparison of the models on real data represented by an MTPL portfolio. The comparison is presented by the results of goodness-of-fit measures. Furthermore, the predictive power of selected models is tested for the given dataset, using the simulation method. Hence, this thesis provides a combination of the analysis of goodness-of-fit results and the predictive power of the models.

In this paper we study the occurrences of outdoor vehicle fires recorded by the Swedish Civil Contingencies Agency (MSB) for the period 1998-2019, and build static panel data models to predict future occurrences of fire in Stockholm County. Through comparing the performance of different models, we look at the effect of different distributional assumptions for the dependent variable on predictive performance. Our study concludes that treating the dependent variable as continuous does not hamper performance, with the exception of models meant to predict more uncommon occurrences of fire. Furthermore, we find that assuming that the dependent variable follows a Negative Binomial Distribution, rather than a Poisson Distribution, does not lead to substantial gains in performance, even in cases of overdispersion. Finally, we notice a slight increase in the number of vehicle fires shown in the data, and reflect on whether this could be related to the increased population size.

No description available.

Les données comptées (count data) possèdent des distributions ayant des caractéristiques particulières comme la non-normalité, l’hétérogénéité des variances ainsi qu’un nombre important de zéros. Il est donc nécessaire d’utiliser les modèles appropriés afin d’obtenir des résultats non biaisés. Ce mémoire compare quatre modèles d’analyse pouvant être utilisés pour les données comptées : le modèle de Poisson, le modèle binomial négatif, le modèle de Poisson avec inflation du zéro et le modèle binomial négatif avec inflation du zéro. À des fins de comparaisons, la prédiction de la proportion du zéro, la confirmation ou l’infirmation des différentes hypothèses ainsi que la prédiction des moyennes furent utilisées afin de déterminer l’adéquation des différents modèles. Pour ce faire, le nombre d’arrestations des membres de gangs de rue sur le territoire de Montréal fut utilisé pour la période de 2005 à 2007. L’échantillon est composé de 470 hommes, âgés de 18 à 59 ans. Au terme des analyses, le modèle le plus adéquat est le modèle binomial négatif puisque celui-ci produit des résultats significatifs, s’adapte bien aux données observées et produit une proportion de zéro très similaire à celle observée. / Count data have distributions with specific characteristics such as non-normality, heterogeneity of variances and a large number of zeros. It is necessary to use appropriate models to obtain unbiased results. This memoir compares four models of analysis that can be used for count data: the Poisson model, the negative binomial model, the Poisson model with zero inflation and the negative binomial model with zero inflation. For purposes of comparison, the prediction of the proportion of zero, the confirmation or refutation of the various assumptions and the prediction of average number of arrrests were used to determine the adequacy of the different models. To do this, the number of arrests of members of street gangs in the Montreal area was used for the period 2005 to 2007. The sample consisted of 470 men, aged 18 to 59 years. After the analysis, the most suitable model is the negative binomial model since it produced significant results, adapts well to the observed data and produces a zero proportion very similar to that observed.

Les données comptées (count data) possèdent des distributions ayant des caractéristiques particulières comme la non-normalité, l’hétérogénéité des variances ainsi qu’un nombre important de zéros. Il est donc nécessaire d’utiliser les modèles appropriés afin d’obtenir des résultats non biaisés. Ce mémoire compare quatre modèles d’analyse pouvant être utilisés pour les données comptées : le modèle de Poisson, le modèle binomial négatif, le modèle de Poisson avec inflation du zéro et le modèle binomial négatif avec inflation du zéro. À des fins de comparaisons, la prédiction de la proportion du zéro, la confirmation ou l’infirmation des différentes hypothèses ainsi que la prédiction des moyennes furent utilisées afin de déterminer l’adéquation des différents modèles. Pour ce faire, le nombre d’arrestations des membres de gangs de rue sur le territoire de Montréal fut utilisé pour la période de 2005 à 2007. L’échantillon est composé de 470 hommes, âgés de 18 à 59 ans. Au terme des analyses, le modèle le plus adéquat est le modèle binomial négatif puisque celui-ci produit des résultats significatifs, s’adapte bien aux données observées et produit une proportion de zéro très similaire à celle observée. / Count data have distributions with specific characteristics such as non-normality, heterogeneity of variances and a large number of zeros. It is necessary to use appropriate models to obtain unbiased results. This memoir compares four models of analysis that can be used for count data: the Poisson model, the negative binomial model, the Poisson model with zero inflation and the negative binomial model with zero inflation. For purposes of comparison, the prediction of the proportion of zero, the confirmation or refutation of the various assumptions and the prediction of average number of arrrests were used to determine the adequacy of the different models. To do this, the number of arrests of members of street gangs in the Montreal area was used for the period 2005 to 2007. The sample consisted of 470 men, aged 18 to 59 years. After the analysis, the most suitable model is the negative binomial model since it produced significant results, adapts well to the observed data and produces a zero proportion very similar to that observed.

Neste trabalho é proposta a família de distribuições Série de Potência k-Modificadas para modelar conjuntos de dados de contagem que apresentam ou não alguma discrepância na frequência da observação k em relação à distribuição Série de Potência associada. É importante ressaltar que o emprego do termo Modificada(s) não possui o mesmo contexto ao empregado por Gupta (1974), o qual introduziu a classe de distribuições Série de Potência Modificadas representada pela sigla MPSD. Neste trabalho, entende-se por modificação, a inclusão de um parâmetro na função massa de probabilidade da distribuição Série de Potência tornando essa nova família de distribuições capaz de modelar adequadamente conjunto de dados para os casos em que há excesso (inflação), falta (deflação), ausência ou até mesmo quando a frequência da observação k estiver de acordo para a suposição de uma distribuição pertencente à família Série de Potência. Para esta nova família de distribuições são apresentadas propriedades como Função de distribuição, Função característica, Função geradora de momentos, Estatísticas de Ordem dentre outras, além de contextualizá-la como modelo de mistura. As distribuições consideradas para a construção dessa nova família serão as distribuições uniparamétricas pertencentes à família Série de Potência, cuja função massa de probabilidade pode ser escrita em função de sua média. / In this work, it is proposed the family of k-modified power series distributions to model count data sets that may or may not present some discrepancy in the frequency of the observation k in relation to the power series distribution associated. It is important to highlight that employing the term \"modified\" does not imply the same context to the one employed by Gupta (1974), which introduced the class of power series modified distributions represented by the acronym MPSD. In this work, modification can be understood as the inclusion of a parameter in the probability mass function of the power series distribution, allowing this family of distributions to properly model a data set for cases where there is an excess (inflation), deficiency (deflation), lack or even when the frequency of observations k are in agreement with the supposition of a distribution belonging to the power series family. It is presented, for this new family of distributions, properties like distribution function, characteristic function, moment generating function, order statistics, among others. Moreover the family is also contextualized as a mixture model. The distributions considered to construct this new family are uniparametric and belong to the power series family, for which the probability mass can be written as function of its mean.

Submitted by Daniele Amaral (daniee_ni@hotmail.com) on 2017-10-10T18:05:40Z No. of bitstreams: 1 DissSOC.pdf: 1290080 bytes, checksum: 3b1d7f1f88c287f6fc22d678e2f968ed (MD5) / Approved for entry into archive by Ronildo Prado (bco.producao.intelectual@gmail.com) on 2018-01-25T12:27:04Z (GMT) No. of bitstreams: 1 DissSOC.pdf: 1290080 bytes, checksum: 3b1d7f1f88c287f6fc22d678e2f968ed (MD5) / Approved for entry into archive by Ronildo Prado (bco.producao.intelectual@gmail.com) on 2018-01-25T12:27:14Z (GMT) No. of bitstreams: 1 DissSOC.pdf: 1290080 bytes, checksum: 3b1d7f1f88c287f6fc22d678e2f968ed (MD5) / Made available in DSpace on 2018-01-25T12:31:09Z (GMT). No. of bitstreams: 1 DissSOC.pdf: 1290080 bytes, checksum: 3b1d7f1f88c287f6fc22d678e2f968ed (MD5) Previous issue date: 2017-05-23 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / This paper proposes a family of distributions Power Series k-Modified from to model sets of count data which have or not any discrepancy in the frequency of observation k in relation to the distribution associated Power Series. Is understood as a modification, inclusion of a parameter in the mass function of probability of the distribution Power Series, making this new family of distributions able to adequately model the data set in cases where there is excess (inflation), poor (deflation) the absence or even when the frequency of the observations k is according to the distribution power series. For this new family of distributions are presented some properties as the distribution functions, Statistics Order among others, besides contextualizes it as mixing model and place it in the context of regression models. The distributions considered for the construction of this new family will be uni-parametric distributions belonging to the Power Series family, whose probability mass function can be written in terms of their average. / Neste trabalho é proposta a família de distribuições Série de Potência k-Modificadas para modelar conjuntos de dados de contagem que apresentam ou não alguma discrepância na frequência da observação k em relação à distribuição Série de Potência associada. É importante ressaltar que o emprego do termo Modificada(s) não possui o mesmo contexto ao empregado por Gupta (1974), o qual introduziu a classe de distribuições Série de Potência Modificadas representada pela sigla MPSD. Neste trabalho, entende-se por modificação, a inclusão de um parâmetro na função massa de probabilidade da distribuição Série de Potência tornando essa nova família de distribuições capaz de modelar adequadamente conjunto de dados para os casos em que há excesso (inflação), falta (deflação), ausência ou até mesmo quando a frequência da observação k estiver de acordo para a suposição de uma distribuição Série de Potência. Para esta nova família de distribuições são apresentadas propriedades como Função de distribuição, Função característica, Função geradora de momentos, Estatística de Ordem dentre outras, além de contextualiza-la como modelo de mistura. As distribuições consideradas para a construção dessa nova família serão as distribuições uniparamétricas pertencentes à família Série de Potência, cuja função massa de probabilidade pode ser escrita em função de sua média.

[pt] Demanda intermitente é definida por períodos de vendas nulas intercaladas com vendas positivas e de quantidade altamente variável. A maior parte das unidades de manutenção de estoque (stock keeping units, em inglês) ao nível loja pode ser caracterizada como contendo demanda desse tipo. Assim, modelos acurados para prever séries com demanda intermitente trazem grandes impactos em relação à gestão de estoque. Nesta dissertação nós propomos o uso do framework GAS com as distribuições adequadas para dados de contagem, além de suas versões com excesso de zeros, e aplicamos os modelos derivados a dados reais obtidos com uma grande rede varejista brasileira. Nós demonstramos que os modelos com excesso de zeros propostos são estimados de forma consistente por máxima verossimilhança e a distribuição dos estimadores é assintóticamente normal. A performance dos modelos propostos é comparada com benchmarks adequados das literaturas de séries temporais para dados de contagem e previsão de demanda intermitente. A avaliação das previsões é feita com base tanto na precisão da distribuição preditiva quanto na precisão das previsões pontuais. Nossos resultados mostram que os modelos propostos, em especial o modelo derivado sob distribuição hurdle Poisson, performam melhor do que os benchmarks analisados. / [en] Intermittent demand is defined by periods of zero sales interleaved with positive sales with highly variable quantities. Most stock keeping units at the store level can be characterized as containing such demand. Thus, accurate models for predicting series with intermittent demand have major impacts in relation to inventory management. In this dissertation we propose the use of the GAS framework with the appropriate distributions for count data, in addition to their versions with excess of zeroes, and apply the derived models to real data obtained from a large Brazilian retail chain. We demonstrate that the proposed models with excess of zeros are consistently estimated via maximum likelihood and the distribution of the estimator is asymptotically normal. The performance of the proposed models is compared to adequate benchmarks from the time series literature for count data and intermittent demand forecast. Forecasting is evaluated based on the accuracy of both the entire predictive distribution and point forecasts. Our results show that the proposed models, specially the one derived from hurdle Poisson distribution, perform better than the analyzed benchmarks.

Ma thèse est composée de trois chapitres reliés à l'estimation des modèles espace-état et volatilité stochastique. Dans le première article, nous développons une procédure de lissage de l'état, avec efficacité computationnelle, dans un modèle espace-état linéaire et gaussien. Nous montrons comment exploiter la structure particulière des modèles espace-état pour tirer les états latents efficacement. Nous analysons l'efficacité computationnelle des méthodes basées sur le filtre de Kalman, l'algorithme facteur de Cholesky et notre nouvelle méthode utilisant le compte d'opérations et d'expériences de calcul. Nous montrons que pour de nombreux cas importants, notre méthode est plus efficace. Les gains sont particulièrement grands pour les cas où la dimension des variables observées est grande ou dans les cas où il faut faire des tirages répétés des états pour les mêmes valeurs de paramètres. Comme application, on considère un modèle multivarié de Poisson avec le temps des intensités variables, lequel est utilisé pour analyser le compte de données des transactions sur les marchés financières. Dans le deuxième chapitre, nous proposons une nouvelle technique pour analyser des modèles multivariés à volatilité stochastique. La méthode proposée est basée sur le tirage efficace de la volatilité de son densité conditionnelle sachant les paramètres et les données. Notre méthodologie s'applique aux modèles avec plusieurs types de dépendance dans la coupe transversale. Nous pouvons modeler des matrices de corrélation conditionnelles variant dans le temps en incorporant des facteurs dans l'équation de rendements, où les facteurs sont des processus de volatilité stochastique indépendants. Nous pouvons incorporer des copules pour permettre la dépendance conditionnelle des rendements sachant la volatilité, permettant avoir différent lois marginaux de Student avec des degrés de liberté spécifiques pour capturer l'hétérogénéité des rendements. On tire la volatilité comme un bloc dans la dimension du temps et un à la fois dans la dimension de la coupe transversale. Nous appliquons la méthode introduite par McCausland (2012) pour obtenir une bonne approximation de la distribution conditionnelle à posteriori de la volatilité d'un rendement sachant les volatilités d'autres rendements, les paramètres et les corrélations dynamiques. Le modèle est évalué en utilisant des données réelles pour dix taux de change. Nous rapportons des résultats pour des modèles univariés de volatilité stochastique et deux modèles multivariés. Dans le troisième chapitre, nous évaluons l'information contribuée par des variations de volatilite réalisée à l'évaluation et prévision de la volatilité quand des prix sont mesurés avec et sans erreur. Nous utilisons de modèles de volatilité stochastique. Nous considérons le point de vue d'un investisseur pour qui la volatilité est une variable latent inconnu et la volatilité réalisée est une quantité d'échantillon qui contient des informations sur lui. Nous employons des méthodes bayésiennes de Monte Carlo par chaîne de Markov pour estimer les modèles, qui permettent la formulation, non seulement des densités a posteriori de la volatilité, mais aussi les densités prédictives de la volatilité future. Nous comparons les prévisions de volatilité et les taux de succès des prévisions qui emploient et n'emploient pas l'information contenue dans la volatilité réalisée. Cette approche se distingue de celles existantes dans la littérature empirique en ce sens que ces dernières se limitent le plus souvent à documenter la capacité de la volatilité réalisée à se prévoir à elle-même. Nous présentons des applications empiriques en utilisant les rendements journaliers des indices et de taux de change. Les différents modèles concurrents sont appliqués à la seconde moitié de 2008, une période marquante dans la récente crise financière. / My thesis consists of three chapters related to the estimation of state space models and stochastic volatility models. In the first chapter we develop a computationally efficient procedure for state smoothing in Gaussian linear state space models. We show how to exploit the special structure of state-space models to draw latent states efficiently. We analyze the computational efficiency of Kalman-filter-based methods, the Cholesky Factor Algorithm, and our new method using counts of operations and computational experiments. We show that for many important cases, our method is most efficient. Gains are particularly large for cases where the dimension of observed variables is large or where one makes repeated draws of states for the same parameter values. We apply our method to a multivariate Poisson model with time-varying intensities, which we use to analyze financial market transaction count data. In the second chapter, we propose a new technique for the analysis of multivariate stochastic volatility models, based on efficient draws of volatility from its conditional posterior distribution. It applies to models with several kinds of cross-sectional dependence. Full VAR coefficient and covariance matrices give cross-sectional volatility dependence. Mean factor structure allows conditional correlations, given states, to vary in time. The conditional return distribution features Student's t marginals, with asset-specific degrees of freedom, and copulas describing cross-sectional dependence. We draw volatility as a block in the time dimension and one-at-a-time in the cross-section. Following McCausland(2012), we use close approximations of the conditional posterior distributions of volatility blocks as Metropolis-Hastings proposal distributions. We illustrate using daily return data for ten currencies. We report results for univariate stochastic volatility models and two multivariate models. In the third chapter, we evaluate the information contributed by (variations of) realized volatility to the estimation and forecasting of volatility when prices are measured with and without error using a stochastic volatility model. We consider the viewpoint of an investor for whom volatility is an unknown latent variable and realized volatility is a sample quantity which contains information about it. We use Bayesian Markov Chain Monte Carlo (MCMC) methods to estimate the models, which allow the formulation of the posterior densities of in-sample volatilities, and the predictive densities of future volatilities. We then compare the volatility forecasts and hit rates from predictions that use and do not use the information contained in realized volatility. This approach is in contrast with most of the empirical realized volatility literature which most often documents the ability of realized volatility to forecast itself. Our empirical applications use daily index returns and foreign exchange during the 2008-2009 financial crisis.