[pt] Nesta dissertação é proposto um modelo híbrido para previsão de carga de curtíssimo prazo, combinando amortecimento exponencial simples e redes neurais artificiais do topo feed-forward. O modelo fornece previsões pontuais e limites superiores e inferiores para um horizonte de quinze dias. Estes limites formam um intervalo ao qual pode ser associado um nível de confiança empírico, estimado através de um teste fora da amostra. O desempenho do modelo é avaliado ao longo de uma simulação realizada com dados reais de duas concessionárias de energia elétrica brasileiras. / [en] This thesis presents an hibrid short term load forecasting model that mixes simple exponential smoothing with feed- forward neural networks. The model gives point predictions with upper and lower limits for 15-day-ahead horizon. These limits yields an interval with associated empirical confidence level, estimated by an out of sample test. The model's performance is evaluated through a simulation with real data obtained from two Brazilian utilities. / [es] En esta disertación se propone un modelo híbrido para previsión de carga de cortísimo plazo, combinando amortecimiento exponencial simple y redes neurales artificiales tipo feed-forward. EL modelo nos da las previsiones puntuales y los límites superiores e inferiores para un horizonte de quince días. Estos límites forman un intervalo al cual se le puede asociar un nível de confianza empírico, estimado a través de un test out of sample. EL desempeño del modelo se evalúa utilizando datos reales de dos concesionarias de energía eléctrica brasileras.

In this thesis, using the principles of confirmatory factor analysis (CFA) and the cause-effect concept associated with structural equation modelling (SEM), a new flexible statistical framework for evaluation of climate model simulations against observational data is suggested. The design of the framework also makes it possible to investigate the magnitude of the influence of different forcings on the temperature as well as to investigate a general causal latent structure of temperature data. In terms of the questions of interest, the framework suggested here can be viewed as a natural extension of the statistical approach of 'optimal fingerprinting', employed in many Detection and Attribution (D&amp;A) studies. Its flexibility means that it can be applied under different circumstances concerning such aspects as the availability of simulated data, the number of forcings in question, the climate-relevant properties of these forcings, and the properties of the climate model under study, in particular, those concerning the reconstructions of forcings and their implementation. It should also be added that although the framework involves the near-surface temperature as a climate variable of interest and focuses on the time period covering approximately the last millennium prior to the industrialisation period, the statistical models, included in the framework, can in principle be generalised to any period in the geological past as soon as simulations and proxy data on any continuous climate variable are available.  Within the confines of this thesis, performance of some CFA- and SEM-models is evaluated in pseudo-proxy experiments, in which the true unobservable temperature series is replaced by temperature data from a selected climate model simulation. The results indicated that depending on the climate model and the region under consideration, the underlying latent structure of temperature data can be of varying complexity, thereby rendering our statistical framework, serving as a basis for a wide range of CFA- and SEM-models, a powerful and flexible tool. Thanks to these properties, its application ultimately may contribute to an increased confidence in the conclusions about the ability of the climate model in question to simulate observed climate changes. / <p>At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 2: Manuscript. Paper 3: Manuscript. Paper 3: Manuscript.</p>

This master´s thesis deals with unconventional technology of wire electrical discharge machining. There are described the principles and essence of electrical discharge machining and the principles related to wire electrical discharge machining with emphasis on the application of this technology in terms of medium-sized engineering company. There is also described the complete assembly of technolgy wire cutting and machining on wire cutter Excetek V 650. Then in the work there are statistically evaluated parameters precision machined surfaces, specifically to the belt pulley 116-8M-130. At the end of the work there is the technical-economic evaluation that addresses the hourly cost of machining on wire cutter Excetek V 650.

Pour construire des intervalles de confiance, nous pouvons utiliser diverses approches bootstrap. Nous avons un problème pour le contexte spécifique d’un paramètre de proportion lorsque l’estimateur usuel, la proportion de succès dans l’échantillon ˆp, est nul. Dans un contexte classique d’observations indépendantes et identiquement distribuées (i.i.d.) de la distribution Bernoulli, les échantillons bootstrap générés ne contiennent que des échecs avec probabilité 1 et les intervalles de confiance bootstrap deviennent dégénérés en un seul point, soit le point 0. En contexte de population finie, nous sommes confrontés aux mêmes problèmes lorsqu’on applique une méthode bootstrap à un échantillon de la population ne contenant que des échecs. Une solution possible s’inspire de l’estimateur utilisé dans les méthodes de [Wilson, 1927] et [Agresti et Coull, 1998] où ceux-ci considèrent ˜p l’estimateur qui prend la proportion de succès d’un échantillon augmenté auquel on a ajouté deux succès et deux échecs. La solution que nous introduisons consiste à effectuer le bootstrap de la distribution de ˆp mais en appliquant les méthodes bootstrap à l’échantillon augmenté de deux succès et deux échecs, tant en statistique classique que pour une population finie. Les résultats ont démontré qu’une version de la méthode percentile est la méthode bootstrap la plus efficace afin d’estimer par intervalle de confiance un paramètre de proportion autant dans un contexte i.i.d. que dans un contexte d’échantillonnage avec le plan aléatoire simple sans remise. Nos simulations ont également démontré que cette méthode percentile pouvait compétitionner avantageusement avec les meilleures méthodes traditionnelles. / A few bootstrap approaches exist to create confidence intervals. Some difficulties appear for the specific case of a proportion when the usual estimator, the proportion of success in a sample, is 0. In the classical case where the observations are independently and identically distributed (i.i.d.) from a Bernoulli distribution, the bootstrap samples only contain zeros with probability 1 and the resulting bootstrap confidence intervals are degenerate at the value 0. We are facing the same problem in the survey sampling case when we apply the bootstrap method to a sample with all observations equal to 0. A possible solution is suggested by the estimator found in the confidence intervals of [Wilson, 1927] and [Agresti et Coull, 1998] where they use ˜p the proportion of success in a augmented sample consisting of adding two successes and two failures to the original sample. The proposed solution is to use the bootstrap method on ˆp but where the bootstrap is based on the augmented sample with two additional successes and failures, whether the sample comes from i.i.d. Bernoulli variables or from a simple random sample. Results show that a version of the percentile method is the most efficient bootstrap method to construct confidence intervals for a proportion both in the classical setting or in the case of a simple random sample. Our results also show that this percentile interval can compete with the best traditional methods.

No description available.

The unique problems encountered when analyzing weather data sets - that is, measurements taken while conducting a meteorological experiment- have forced statisticians to reconsider the conventional analysis methods and investigate permutation test procedures. The problems encountered when analyzing weather data sets are simulated for a Monte Carlo study, and the results of the parametric and permutation t-tests are compared with regard to significance level, power, and the average coilfidence interval length. Seven population distributions are considered - three are variations of the normal distribution, and the others the gamma, the lognormal, the rectangular and empirical distributions. The normal distribution contaminated with zero measurements is also simulated. In those simulated situations in which the variances are unequal, the permutation test procedure was performed using other test statistics, namely the Scheffe, Welch and Behrens-Fisher test statistics. / Mathematical Sciences / M. Sc. (Statistics)

Ce mémoire porte sur la simulation d'intervalles de crédibilité simultanés dans un contexte bayésien. Dans un premier temps, nous nous intéresserons à des données de précipitations et des fonctions basées sur ces données : la fonction de répartition empirique et la période de retour, une fonction non linéaire de la fonction de répartition. Nous exposerons différentes méthodes déjà connues pour obtenir des intervalles de confiance simultanés sur ces fonctions à l'aide d'une base polynomiale et nous présenterons une méthode de simulation d'intervalles de crédibilité simultanés. Nous nous placerons ensuite dans un contexte bayésien en explorant différents modèles de densité a priori. Pour le modèle le plus complexe, nous aurons besoin d'utiliser la simulation Monte-Carlo pour obtenir les intervalles de crédibilité simultanés a posteriori. Finalement, nous utiliserons une base non linéaire faisant appel à la transformation angulaire et aux splines monotones pour obtenir un intervalle de crédibilité simultané valide pour la période de retour. / This master's thesis addresses the problem of the simulation of simultaneous credible intervals in a Bayesian context. First, we will study precipation data and two functions based on these data : the empirical distribution function and the return period, a non-linear function of the empirical distribution. We will review different methods already known to obtain simultaneous confidence intervals of these functions with a polynomial basis and we will present a method to simulate simultaneous credible intervals. Second, we will explore some models of prior distributions and in the more complex one, we will need the Monte-Carlo method to simulate simultaneous posterior credible intervals. Finally, we will use a non-linear basis based on the angular transformation and on monotone splines to obtain valid simultaneous credible intervals for the return period.

No description available.

[pt] Em obras de maior porte e grande responsabilidade (portos, barragens, usinas nucleares, etc.), a quantidade de instrumentações pode se tornar suficiente para permitir a construção de modelos de variabilidade temporal das propriedades de interesse com base em redes neurais artificiais. No caso de barragens, o monitoramento através da instalação de um sistema de instrumentação desempenha um papel fundamental na avaliação do comportamento destas estruturas, tanto durante o período de construção quanto no período de operação. Neste trabalho empregou-se a técnica de redes neurais temporais (RNT) para análise, modelagem e previsão dos valores de vazão na barragem Funil, do sistema Furnas Centrais Elétricas, a partir dos dados de instrumentações disponíveis no período compreendido entre 02/09/1985 e 25/02/2002. As redes neurais temporais empregadas foram: RNT com arquitetura feedforward associada a técnica de janelamento, RNT recorrente Elman, RNT FIR e RNT Jordan. Adicionalmente, foram utilizadas duas técnicas para análise das séries temporais: os modelos de Box & Jenkins (1970) e métodos geoestatísticos, com a finalidade de comparar com o desempenho das RNT´s. Nesta pesquisa estuda-se ainda a geração de intervalos de confiança para RNT e para métodos geoestatísticos. As previsões de vazão analisadas neste trabalho, envolvendo o comportamento da barragem Funil, apresentaram resultados satisfatórios tanto os obtidos pelos modelos de redes neurais temporais como pelos de Box & Jenkins e métodos geoestatísticos. / [en] In works of great responsibility (ports, dams, nuclear power, etc.), the amount of instrumentation data may allow the construction of models for the temporary variability of the properties of interest based on neural network techniques. In case of dams, the monitoring through the installation of an instrumentation system plays a fundamental part in the evaluation of the behavior of these structures, during the construction period as well as in the operation period. In this work the technique of temporal neural networks (TNN) was used for analysis, modeling and forecast of the water discharges values in the Funil dam, from Furnas Centrais Elétricas system, starting from the data of available instrumentation in the period between 02/09/1985 and 25/02/2002. The temporal neural networks used in this research were the following: TNN with feedforward architecture and the windowing technique, recursive TNN Elman, TNN FIR and TNN Jordan. Two additional techniques (Box & Jenkins and geostatistical models) were employed for analysis of the time series with the purpose to compare the results obtained with neural networks. In this research the generation of confidence intervals for TNN and geostatistical methods were also investigated. The discharge values forecasts analyzed in this work for the Funil dam presented satisfactory results, with respect to the neural network, Box & Jenkins and geostatistical methods.

The unique problems encountered when analyzing weather data sets - that is, measurements taken while conducting a meteorological experiment- have forced statisticians to reconsider the conventional analysis methods and investigate permutation test procedures. The problems encountered when analyzing weather data sets are simulated for a Monte Carlo study, and the results of the parametric and permutation t-tests are compared with regard to significance level, power, and the average coilfidence interval length. Seven population distributions are considered - three are variations of the normal distribution, and the others the gamma, the lognormal, the rectangular and empirical distributions. The normal distribution contaminated with zero measurements is also simulated. In those simulated situations in which the variances are unequal, the permutation test procedure was performed using other test statistics, namely the Scheffe, Welch and Behrens-Fisher test statistics. / Mathematical Sciences / M. Sc. (Statistics)