Les espèces démersales représentent 50% des captures des pêcheries françaises du golfe du Lion, dont la plupart sont pleinement exploitées, voir surexploitées depuis plusieurs décennies. Cette thèse évalue la pertinence d'aires marines protégées (AMPs) comme outil de gestion et de conservation de ces populations. Jusqu'à présent de telles zones ont été uniquement mises en place le long des côtes pour protéger des espèces très peu mobiles. Le problème est plus complexe pour les espèces vivant en haute mer car leur habitat est plus large et plus diffus. Pour ce faire, la distribution spatiale de 12 espèces démersales exploitées clés ont été étudiées de 1994 à 2010, à l'aide d'observations scientifiques et d'outils statistiques ad hoc. Une approche géostatistique a permis de détecter des structures d'auto-corrélation spatiale pour l'ensemble des espèces, et de produire des cartes de distributions annuelles de chaque espèce. Ces distributions sont apparues très stables sur les dix-sept années, mis à part un phénomène d'expansion/ rétraction avec le niveau d'abondance totale sur la région. Par ailleurs une approche par modèle linéaire généralisé a révélé des associations importantes de ces espèces à un habitat stable dans le temps. Ces résultats sont en accord avec la théorie du bassin de MacCall selon laquelle l'association d'une espèce à un habitat est densité-dépendant, et l'augmentation de la densité d'individus dans une zone serait à l'origine de la colonisation d'habitats sub-optimaux. Protéger l'habitat optimal d'une espèce permettrait alors de constituer un habitat «source», si la zone est judicieusement choisie ; en effet le report de l'effort de pêche à l'extérieur de l'AMP peut au contraire rendre cette mesure inefficace voir délétère. Par ailleurs les populations adultes occupaient généralement des zones plus concentrées et incluses dans l'aire de répartition des juvéniles. Ces zones communes d'habitats essentiels (reproduction et nourriceries) peuvent être potentiellement intéressantes à protéger dans un cadre monospécifique. L'hétérogénéité observée des répartitions d'une espèce à l'autre implique l'instauration de zones très clairsemées, et rend la gestion difficile dans une cadre pluri-spécifique. Une zone de taille raisonnable a tout de même été identifiée, représentant 20% de la population de chaque espèce et représentative de la diversité des habitats de cette région. / Demersal species represent 50% of French fisheries catches in theGulf of Lions, most of which are fully exploited, or overfished for decades. This thesis evaluates the relevance of marine protected areas (MPAs) as a tool for conservation and management of these populations. So far these areas have been implemented only along the coast to protect the very few mobile species. The problem is more complex for deep sea species because their habitat is broader and more diffuse. To do this, the spatial distribution of 12 key demersal species exploited were studied from 1994 to 2010, with scientific observations and ad hoc statistical tools. A geostatistical approach allowed to detect spatial autocorrelation structures for all species, and produce maps of annual distributions of each species. These distributions appeared very stable over 17 years, apart from a phenomenon of expansion/ contraction with the level of total abundance in the region. In addition, a generalized linear model approach revealed significant associations of these species to a temporally stable habitat. These results are consistent with MacCall basin theory, according which habitat suitability is a density-dependent thus the increase of individuals in an area make them colonize sub-optimal habitats. An optimal habitat under protection could thus be "source" habitat, if the area is carefully chosen. Indeed reporting the fishing effort outside the MPA can instead make this measure ineffective or deleterious. The adult population were generally in more concentrated areas and included in the spatial range of juveniles. These common areas of essential habitat (breeding and nursery) may be potentially interesting to protect a single species . However, the heterogeneity of distributions of a species to another involves the introduction of very sparse areas, making the management difficult. However an area of ​​reasonable size has been identified, covering 20% ​​of the population of each species and representative of the diversity of bottom habitats in the region.

O presente trabalho tem por objetivo utilizar a abordagem dos modelos lineares generalizados e os modelos de dispersão no contexto do seguro agrícola. Os modelos lineares generalizados (MLG\'s) constituem uma extensão dos modelos lineares de regressão múltipla introduzida por Nelder e Wedderburn (1972), que inclui modelos cuja variável resposta pertence à família exponencial de distribuições. O MLG é formado por um componente aleatório, que possui distribuição pertencente à família exponencial, um componente sistemático, conectados por uma função de ligação. Jorgensen (1997) estende a utilização dos MLG para uma classe mais ampla de modelos probabilísticos, denominados modelos de dispersão. A estimação dos parâmetros foi baseada no método da máxima verossimilhança, e também, em função da amostra ser relativamente pequena, optou-se pelo método de bootstrap não-paramétrico. As duas abordagens foram aplicadas a dois conjuntos de dados de sinistros de 15 municípios do estado do Rio Grande do Sul. Os resultados mostraram que a precipitação acumulada tem influência na ocorrência de sinistros. Entretanto, na modelagem do montante do sinistro não foi encontrada nenhuma variável significativa. Usando o método de bootstrap, foi encontrada influência das variáveis precipitação acumulada e a temperatura média no numero de sinistros / The main objective of this work is to use the generalized linear models and dispersion models in the agricultural insurance context. The Generalized Linear Model (GLM) are an extension of the multiple regression linear models presented by Nelder e Wedderburn (1972). This approach include situations in which the response variable can be included in exponencial the family. The GLM is composed of a randomized component, a sistematic component and the link functions. JÁrgensen (1997) extend the application of the GLM for a more general class of probability models, called dispersion models. Both approaches were applied in two insurance datasets for 15 citys in Rio Grande do Sul. The parameters estimation was based in the maximum likelihood method, in addition, because of the relatively small sample, the non-parametric Bootstrap method was used. This study show, using GLM, that only the accumulated rainfall was statistically significant . However, any of the covariates was significant when modelling the amount of claims. In the analysis using Bootstrap method the accumulated rainfall and average temperature were significant when modelling the number of insurance clains.

The objective of this research is to evaluate and introduce a new methodology regarding rural highway safety. Current practices rely on crash prediction models that utilize specific explanatory variables, whereas the depository of knowledge for past research is the Highway Safety Manual (HSM). Most of the prediction models in the HSM identify the effect of individual geometric elements on crash occurrence and consider their combination in a multiplicative manner, where each effect is multiplied with others to determine their combined influence. The concepts of 3-dimesnional (3-D) representation of the roadway surface have also been explored in the past aiming to model the highway structure and optimize the roadway alignment. The use of differential geometry on utilizing the 3-D roadway surface in order to understand how new metrics can be used to identify and express roadway geometric elements has been recently utilized and indicated that this may be a new approach in representing the combined effects of all geometry features into single variables. This research will further explore this potential and examine the possibility to utilize 3-D differential geometry in representing the roadway surface and utilize its associated metrics to consider the combined effect of roadway features on crashes. It is anticipated that a series of single metrics could be used that would combine horizontal and vertical alignment features and eventually predict roadway crashes in a more robust manner. It should be also noted that that the main purpose of this research is not to simply suggest predictive crash models, but to prove in a statistically concrete manner that 3-D metrics of differential geometry, e.g. Gaussian Curvature and Mean Curvature can assist in analyzing highway design and safety. Therefore, the value of this research is oriented towards the proof of concept of the link between 3-D geometry in highway design and safety. This thesis presents the steps and rationale of the procedure that is followed in order to complete the proposed research. Finally, the results of the suggested methodology are compared with the ones that would be derived from the, state-of-the-art, Interactive Highway Safety Design Model (IHSDM), which is essentially the software that is currently used and based on the findings of the HSM.

This thesis studies the Second-order Least Squares (SLS) estimation method in regression models with and without measurement error. Applications of the methodology in general quasi-likelihood and variance function models, censored models, and linear and generalized linear models are examined and strong consistency and asymptotic normality are established. To overcome the numerical difficulties of minimizing an objective function that involves multiple integrals, a simulation-based SLS estimator is used and its asymptotic properties are studied. Finite sample performances of the estimators in all of the studied models are investigated through simulation studies. / February 2009

The main topic of the paper is on-line filtering for non-Gaussian dynamic (state space) models by approximate computation of the first two posterior moments using efficient numerical integration. Based on approximating the prior of the state vector by a normal density, we prove that the posterior moments of the state vector are related to the posterior moments of the linear predictor in a simple way. For the linear predictor Gauss-Hermite integration is carried out with automatic reparametrization based on an approximate posterior mode filter. We illustrate how further topics in applied state space modelling such as estimating hyperparameters, computing model likelihoods and predictive residuals, are managed by integration-based Kalman-filtering. The methodology derived in the paper is applied to on-line monitoring of ecological time series and filtering for small count data. (author's abstract) / Series: Forschungsberichte / Institut für Statistik

The topic of the paper is filtering for non-Gaussian dynamic (state space) models by approximate computation of posterior moments using numerical integration. A Gauss-Hermite procedure is implemented based on the approximate posterior mode estimator and curvature recently proposed in 121. This integration-based filtering method will be illustrated by a dynamic trend model for non-Gaussian time series. Comparision of the proposed method with other approximations ([15], [2]) is carried out by simulation experiments for time series from Poisson, exponential and Gamma distributions. (author's abstract) / Series: Forschungsberichte / Institut für Statistik

The aim of the monograph is to contribute towards bridging the gap between methodological developments that have evolved in the social sciences, in particular in psychometric research, and methods of statistical modelling in a more general framework. The first part surveys certain special psychometric models (often referred to as Rasch family of models) that share common properties: separation of parameters describing qualities of the subject under investigation and parameters related to properties of the situation under which the response of a subject is observed. Using conditional maximum likelihood estimation, both types of parameters may be estimated independently from each other. In particular, the Rasch model, the rating scale model, the partial credit model, hybrid types, and linear extensions thereof are treated. The second part reviews basic ideas of generalized linear models (GLMs) as an an excellent framework for unifying different approaches and providing a natural, technical background for model formulation, estimation and testing. This is followed by a short introduction to the software package GLIM chosen to illustrate the formulation of psychometric models in the GLM framework. The third part is the main part of this monograph and shows the application of generalized linear models to psychometric approaches. It gives a unified treatment of Rasch family models in the context of log-linear models and contains some new material on log-linear longitudinal modelling. The last part of the monograph is devoted to show the usefulness of the latent variable approach in a variety of applications, such as panel, cross-over, and therapy evaluation studies, where standard statistical analysis does not necessarily lead to satisfactory results. (author´s abstract) / Series: Research Report Series / Department of Statistics and Mathematics

This thesis studies the Second-order Least Squares (SLS) estimation method in regression models with and without measurement error. Applications of the methodology in general quasi-likelihood and variance function models, censored models, and linear and generalized linear models are examined and strong consistency and asymptotic normality are established. To overcome the numerical difficulties of minimizing an objective function that involves multiple integrals, a simulation-based SLS estimator is used and its asymptotic properties are studied. Finite sample performances of the estimators in all of the studied models are investigated through simulation studies.

This thesis studies the Second-order Least Squares (SLS) estimation method in regression models with and without measurement error. Applications of the methodology in general quasi-likelihood and variance function models, censored models, and linear and generalized linear models are examined and strong consistency and asymptotic normality are established. To overcome the numerical difficulties of minimizing an objective function that involves multiple integrals, a simulation-based SLS estimator is used and its asymptotic properties are studied. Finite sample performances of the estimators in all of the studied models are investigated through simulation studies.

Submitted by Leonardo Cavalcante (leo.ocavalcante@gmail.com) on 2018-06-06T12:02:25Z No. of bitstreams: 1 Arquivototal.pdf: 3264478 bytes, checksum: 742da63cecf21ce169225aebd5fea82e (MD5) / Made available in DSpace on 2018-06-06T12:02:25Z (GMT). No. of bitstreams: 1 Arquivototal.pdf: 3264478 bytes, checksum: 742da63cecf21ce169225aebd5fea82e (MD5) Previous issue date: 2016-04-07 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / Introduction: The global warming has severe environmental impacts and reach different areas and localities, in particular the least developed countries in the tropical region. In Brazil, the regions that suffer most impacts, with such changes will be the Amazon and Northeast. With the development of modern society, it is observed that individuals are consuming much of their time indoors, so, numerous studies have sought to study environmental comfort conditions to assist in the execution of activities and maximize human performance. One of the many changes that took place is in the learning environment, which is different compared to years ago. Technological resources in the classroom have made this a more versatile environment, allowing students and teachers to interact with other institutions, providing greater flexibility in the use of tools that extend their cognitive abilities. However, these tools can both facilitate communication and versatility of teaching and learning as increase the thermal load in the environment that already suffers direct influence of the external environment. In this sense, important to investigate how it is the variables of thermal comfort affect students' performance in climate-controlled learning environments provided these technologies. Objective: We analyzed the relationship between thermal comfort and performance of students in environments with technological innovations in areas of northeastern Brazil. Materials and Methods: Data were obtained during the experiments in the two institutions classrooms (A and B) of higher education located in Northeast Brazil. The Institution A is a private university in the city of João Pessoa- PB and the institution B is a public university in city of Teresina-PI. A total of 103 students of the engineering course participated, with one session per day, per class, on three consecutive days. The air temperature through the air conditioning was adjusted to 20ºC and 24ºC; the last day and is turned off the device to try to reach the value of 30 ° C. By heat stress meter TGD400 and microclimate station Babuc A, positioned in the center of the room recorded the parameters of the thermal environment. Cognitive abilities were assessed by the five events that make up the BPR-5 instrument, broken down for the three days of the experiment. Subjective and personal variables were evaluated in accordance with ISO7730/2005, through check list (thermal resistance of clothing) and scales of 4 points (Comfort) and 7 points (Sensation and Desire). With the results and data obtained were carried out simulations using the Generalized Linear Models (GLM) to verify the relationship between thermal comfort and performance. Results: At the institution A 42,9% of the first day of collection, where the temperature was in it 21,49ºC, indicated to be feeling slightly warm environment (+1) and 51,9% said that the place was comfortable (1). The PMV index of ISO7730/2005 indicated that the room was cold, with 57% of the vote between -1,5 and -2,5. On the second day, with an average temperature was 23,24ºC, about 58,1% of the students felt the thermally neutral environment (0) and 55,4% rated it as comfortable (1). The PMV index, on the second day, indicated that the room was slightly cold, with 87,8% of the vote between -0,5 and -1,5. On the third day the mean temperature was 27,49ºC and 41,9% of the rated as hot environment (+2) and finally, 35.1% VII of the evaluated local rated as slightly uncomfortable (2). The Middle Foretold vote indicated that the room was slightly hot on the third day, with 60.8% of votes ranging from +0,5 to +1,5. The institution B, on the first day of collection whose average temperature was 20,07ºC, about 50% of students reported feeling that the environment was slightly cold (-1) and 61.5% of students rated the environment as comfortable (1). The PMV index indicated that the environment was cold with 92% of the vote results were between -1,5 and -2,5. On the second day the average temperature was 33,72ºC and 59,3% of students said they felt the warm environment (+2) and 66,7% rated it as uncomfortable (3). Regarding the PMV 100% of the frequency of the votes indicated that the environment was rated as very hot (PMV> +2.5). On the third day with the average temperature 22,95ºC, about 73,3% of the students indicated that the thermal sensation was neutral (0) and the place was rated by 96,2% of the students as comfortable (1). The PMV index showed that 88,5% of the calculated votes of the students indicated that the site was slightly cold (-0,5 <PMV <-1,5). Using Generalized Linear Models, in particular the Ordinal Regression, was observed from the data of the two institutions, the performance is linked to subjective criteria, in particular the thermal sensation and the globe temperature. It was observed that the rise globe temperature in 1 ° C increases the chance of students pass a performance range lower to a higher 34% for situations where the wind chill is indicated as very hot. Final considerations: This study showed that there is a difference between what the standard ISO7730/2005 predicts what is actually reported by students in areas of the Brazilian Northeast, ie situations where the standard indicates discomfort students report being comfortable. It was also observed that subjective variables are important factors for a good cognitive performance of students. / Introdução: O aquecimento global recente tem impactos ambientais intensos e atingirá diversas áreas e localidades, em especial, os países menos desenvolvidos situados na região tropical. No Brasil, as regiões que mais sofrerão impactos, com tais mudanças, serão a Amazônia e o Nordeste. Com o desenvolvimento da sociedade moderna, observa-se que os indivíduos estão consumindo grande parte do seu tempo em ambientes fechados, e por isso, numerosas pesquisas vêm buscando estudar condições de conforto ambiental que auxiliem na execução das atividades e maximização do desempenho humano. Uma das diversas mudanças que ocorreram, encontra-se no ambiente de ensino, o qual é diferente quando comparado ao de anos atrás. Recursos tecnológicos em sala de aula têm tornado este ambiente mais versátil, permitindo que alunos e professores interajam com outras instituições, conferindo maior flexibilidade na utilização de ferramentas que estendem suas habilidades cognitivas. Nesse sentido, torna-se importante investigar como as variáveis do conforto térmico afetam o desempenho dos alunos em ambientes de ensino climatizados providos destas tecnologias. Objetivo: Analisar a relação entre conforto térmico e o desempenho de estudantes nos ambientes com inovações tecnológicas em áreas do nordeste brasileiro. Materiais e Métodos: Os dados foram obtidos durante a realização dos experimentos nas salas de aula de duas instituições (A e B) de ensino superior localizadas no Nordeste brasileiro. A instituição A é uma universidade privada e encontra-se na cidade de João Pessoa- e a instituição B é uma universidade publica e encontra-se na cidade de Teresina. No total de 103 alunos dos cursos de exatas das duas instituições participaram, sendo uma sessão por dia, por turma, em três dias consecutivos. A temperatura do ar, através do ar condicionado, foi ajustada para 20ºC e 24ºC; no último dia desligou-se o aparelho para tentar se chegar ao valor de 30°C. Através do medidor de estresse térmico TGD400 e estação microclimática Babuc A,posicionados no centro da sala, registraram-se os parâmetros do ambiente térmico. As habilidades cognitivas foram avaliadas pelas cinco provas que compõem o instrumento do BPR-5, decomposto para os três dias de experimento. As variáveis subjetivas e pessoais foram avaliadas de acordo com ISO7730/2005, através de check list (resistência térmica das vestimentas) e escalas de 4 pontos (Conforto) e 7 pontos (Sensação e Desejo). Com os resultados e dados obtidos realizaram-se simulações através dos Modelos Lineares Generalizados (MLG) para se verificar a relação entre conforto térmico e desempenho. Resultados: Na instituição A 42,9% dos alunos no primeiro dia de coleta, onde a temperatura encontrava-se a 21,49ºC, indicaram estar sentindo o ambiente levemente quente (+1) e 51,9% avaliaram que o local estava confortável (1). O índice PMV da ISO7730/2005 indicou que o ambiente estava frio, com 57% dos votos entre -1,5 e -2,5. No segundo dia, cuja temperatura média foi de 23,24ºC, cerca de 58,1% dos alunos sentiram o ambiente termicamente neutro (0) e 55,4% classificaram-no como confortável (1). O índice PMV, no segundo dia, indicou que o ambiente estava levemente frio, com 87,8% dos votos entre -0,5 e -1,5. No terceiro dia a temperatura média foi de 27,49ºC e 41,9% dos alunos classificaram o ambiente como quente (+2) e por fim, 35,1% dos avaliados classificaram o local como levemente desconfortável (2). V O Voto Médio Predito indicou que o ambiente estava levemente quente no terceiro dia, com 60,8% dos votos variando de +0,5 a +1,5. Na instituição B, no primeiro dia de coleta cuja temperatura média foi de 20,07ºC, cerca de 50% dos estudantes indicaram sentir que o ambiente estava levemente frio (-1) e 61,5% dos alunos classificaram o ambiente como confortável (1). O índice PMV indicou que o ambiente estava frio com 92% dos resultados obtidos dos votos estavam entre -1,5 a -2,5. No segundo dia a temperatura média foi de 33,72ºC e 59,3% dos alunos afirmaram que sentiam o ambiente quente (+2) e 66,7% classificaram-no como desconfortável (3). Em relação ao PMV 100% da frequência dos votos indicaram que o ambiente foi classificado como muito quente (PMV> +2,5). No terceiro dia com a temperatura média de 22,95ºC, cerca de 73,3% dos estudantes indicaram que a sensação térmica era neutra (0) e o local foi classificado por 96,2% dos alunos como confortável (1). O índice PMV mostrou que 88,5% dos votos calculados dos alunos, indicaram que o local estava levemente frio (-0,5<PMV<-1,5). Utilizando Modelos Lineares Generalizados, em especial a Regressão Linear Ordinal, observou-se através dos dados das duas instituições, que o desempenho está ligado aos parâmetros subjetivos, em especial a sensação térmica e a temperatura de globo. Foi observado que ao se elevar a temperatura de globo em 1ºC aumenta-se a chance dos estudantes passarem de uma faixa de desempenho inferior para uma superior em 34%, para situações onde a sensação térmica é indicada como muito quente. Considerações finais: Este estudo mostrou que existe uma diferença entre o que a norma ISO7730/2005 prediz com o que realmente é relatado pelos estudantes, de áreas da região do nordeste brasileiro, ou seja, situações em que a norma indica desconforto os estudantes relatam estar confortáveis. Observou-se também que as variáveis subjetivas são fatores importantes para um bom desempenho cognitivo dos estudantes.