Many statistical models, like measurement error models, a general class of survival models, and a mixture data model with random censoring, are semiparametric where interest lies in estimating finite-dimensional parameters in the presence of infinite-dimensional nuisance parameters. Developing efficient estimators for the parameters of interest in these models is important because such estimators provide better inferences. For a general regression model with measurement error, we utilize semiparametric theory to develop an unprecedented estimation procedure which delivers consistent estimators even when the model error and latent variable distributions are misspecified. Until now, root-$n$ consistent estimators for this setting were not attainable except for special cases, like a polynomial relationship between the response and mismeasured variables. Through simulation studies and a nutrition study application, we demonstrate that our method outperforms existing methods which ignore measurement error or require a correct model error distribution. In randomized clinical trials, scientists often compare two-sample survival data with a log-rank test. The two groups typically have nonproportional hazards, however, and using a log rank test results in substantial power loss. To ameliorate this issue and improve model efficiency, we propose a model-free strategy of incorporating auxiliary covariates in a general class of survival models. Our approach produces an unbiased, asymptotically normal estimator with significant efficiency gains over current methods. Lastly, we apply semiparametric theory to mixture data models common in kin-cohort designs of Huntington's disease where interest lies in comparing the estimated age-at-death distributions for disease gene carriers and non-carriers. The distribution of the observed, possibly censored, outcome is a mixture of the genotype-specific distributions where the mixing proportions are computed based on the genotypes which are independent of the trait outcomes. Current methods for such data include a Cox proportional hazards model which is susceptible to model misspecification, and two types of nonparametric maximum likelihood estimators which are either inefficient or inconsistent. Using semiparametric theory, we propose an inverse probability weighting estimator (IPW), a nonparametrically imputed estimator and an optimal augmented IPW estimator which provide more reasonable estimates for the age-at-death distributions, and are not susceptible to model misspecification nor poor efficiencies.

The thesis deals with the theoretical and practical aspects of reinsurance. In the theoretical part the types and forms of reinsurance cover are described including the advantages and disadvantages for each one of them. The thesis also describes the reinsurance from the legal point of view and the explanation of contract clauses is included. Special accounting treatment of reinsurance is described and at the end is the practical part which deals with mapping of processes in the reinsurance department.

Indiana University-Purdue University Indianapolis (IUPUI) / The crossing of Kaplan-Meier survival curves presents a challenge when conducting survival analysis studies, making it unclear whether any of the study groups involved present any significant difference in survival. An approach involving the determination of maximum vertical distance between the curves is considered here as a method to assess whether a survival advantage exists between different groups of patients. The method is illustrated on a dataset containing survival times of patients treated with two cancer treatment regimes, one involving treatment by chemotherapy alone, and the other by treatment with both chemotherapy and radiotherapy.

The main topic of my master thesis is the economy of facultative reinsurance from the point of view of a reinsurance company. First, the thesis briefly deals with the general structure of reinsurance and then is focused only on the facultative part. The thesis puts emphasis on the specific characteristics of facultative reinsurance and the creation of reinsurance slip and then uses this knowledge in a case study. The case study shows the risk from the insurance company and reinsurance company side and its aim is to create proper reinsurance structure for a power plant in Bulgaria. The thesis finishes with a chapter about the development of several reinsurance companies in the market.

No description available.

Made available in DSpace on 2015-03-26T13:32:17Z (GMT). No. of bitstreams: 1 texto completo.pdf: 764829 bytes, checksum: 8dbd03463c4800428f75900ca1340eb0 (MD5) Previous issue date: 2012-07-23 / The use of logistic regression modeling as a tool for modeling statistical probability of an event as a function of one or more independents variables, has grown among researchers in several areas, including Phytopathology. At about the dichotomous logistic regression in which the dependent variable is the type binary or dummy, is the extensive number of studies in the literature that discuss the modeling assumptions and the interpretation of the analyzes, as well as alternatives for implementation in statistical packages. However, when the variable response requires the use three or more categories, the number of publications is scarce. This is not only due to the scarcity of relevant publications on the subject, but also the inherent difficulty of coverage on the subject. In this paper we address the applicability of the model polytomous ordinal logistic regression, as well as differences between the proportional odds models, nonproportional and partial proportional odds. For this, we analyzed data from an experiment in which we evaluated the potential antagonistic fungus Clonostachys rosea in biocontrol of the disease called "gray mold", caused by Botrytis cinerea in strawberry and tomato. The partial proportional odds models and nonproportional were adjusted and compared, since the proportionality test score accused rejection of the proportional odds assumption. The estimates of the model coefficients as well as the odds ratios were interpreted in practical terms for Phytopathology. The polytomous ordinal logistic regression is introduced as an important statistical tool for predicting values, showing the potential of C. rosea in becoming a commercial product to be developed and used in the biological control of the disease, because the application of C. rosea was as or more effective than the use of fungicides in the control of gray mold. / O uso da regressão logística como uma ferramenta estatística para modelar a probabilidade de um evento em função de uma ou mais variáveis explicativas, tem crescido entre pesquisadores em várias áreas, inclusive na Fitopatologia. À respeito da regressão logística dicotômica, na qual a variável resposta é do tipo binária ou dummy, é extenso o número de trabalhos na literatura que abordam a modelagem, as pressuposições e a interpretação das análises, bem como alternativas de implementação em pacotes estatísticos. No entanto, quando a variável resposta requer que se utilize três ou mais categorias, o número de publicações é escasso. Isso devido não somente à escassez de publicações relevantes sobre o assunto, mas também à inerente dificuldade de abrangência sobre o tema. No presente trabalho aborda-se a aplicabilidade do modelo de regressão logística politômica ordinal, bem como as diferenças entre os modelos de chances proporcionais, chances proporcionais parciais e chances não proporcionais. Para isso, foram analisados dados de um experimento em que se avaliou o potencial do fungo antagonista Clonostachys rosea no biocontrole da doença denominada mofo cinzento , causada por Botrytis cinerea em morangueiro e tomateiro. Os modelos de chances proporcionais parciais e não proporcionais foram ajustados e comparados, uma vez que o teste score de proporcionalidade acusou rejeição da pressuposição de chances proporcionais. As estimativas dos coeficientes dos modelos bem como das razões de chances foram interpretadas em termos práticos para a Fitopatologia. A regressão logística politômica ordinal se apresentou como uma importante ferramenta estatística para predição de valores, mostrando o potencial do C. rosea em se tornar um produto comercial a ser desenvolvido e usado no controle biológico da doença, pois a aplicação de C. rosea foi tão ou mais eficiente do que a utilização de fungicidas no controle do mofo cinzento.

Cette thèse présente une approche globale d’optimisation vis-à-vis de la fatigue des matériaux et structures en alliages à mémoire de forme (AMF). Cette approche s’articule en trois étapes : i) Le développement d’une loi de comportement capable de prédire la réponse thermomécanique à l’état stabilisé d’une structure en AMF sous chargement cyclique multiaxial non proportionnel. On prend notamment en compte la dépendance de la déformation résiduelle par rapport à la température. Par ailleurs, la méthode LATIN à grand incrément de temps a été généralisée pour les AMF dans le cadre du modèle ZM. Ceci permet de résoudre les problèmes de convergence numérique rencontrés lorsque le processus de transformation de phase se produit avec une pente du plateau de transformation faible. ii) Le développement d’un critère de fatigue à grand nombre de cycles pour les AMF. Ce critère s’inscrit dans le cadre de la théorie d’adaptation à l’instar du critère de Dang Van pour les métaux élasto-plastiques. Le critère proposé permet de calculer en chaque point de la structure en AMF un facteur de fatigue indiquant son degré de dangerosité. iii) Le développement d’une approche d’optimisation structurale qui peut être utilisée pour améliorer la durée de vie en fatigue prédite par le critère proposé dans la deuxième partie. Des exemples numériques sont traités pour valider chaque étape. L‘approche globale a par ailleurs été testée et validée pour l’optimisation structurale d’un stent. / This thesis presents a comprehensive and effi cient structural optimization approach for shape memory alloys (SMAs) with respect to fatigue. The approach consists of three steps: First, the development of a suitable constitutive model capable of predicting, with good accuracy, the stabilized thermomechanical stress state of a SMA structure subjected to multiaxial nonproportional cyclic loading. The dependence of the saturated residual strain on temperature and loading rate is discussed. In order to overcome numerical convergence problems in situations where the phase transformation process presents little or no positivehardening, the large time increment method (LATIN) is utilized in combination with the ZM (Zaki-Moumni) model to simulate SMA structures instead of conventional incremental methods. Second, a shakedown-based fatigue criterion analogous to the Dang Van model for elastoplastic metals is derived for SMAs to predict whether a SMA structure subjected to high-cycle loading would undergo fatigue. The proposed criterion computes a fatigue factor at each material point, indicating its degree of safeness with respect to high-cycle fatigue. Third, a structural optimization approach, which can be used to improve the fatigue lifetime estimated using the proposed fatigue criterion is presented. The prospects of this work include the validation of the optimization approach with experimental data.