Neste trabalho foi proposto um modelo de regressão para dados com censura intervalar utilizando a distribuição Weibull-exponenciada, que possui como característica principal a função de taxa de falha que assume diferentes formas (unimodal, forma de banheira, crescente e decrescente). O atrativo desse modelo de regressão é a sua utilização para discriminar modelos, uma vez que o mesmo possui como casos particulares os modelos de regressão Exponencial, Weibull, Exponencial-exponenciada, entre outros. Também foi estudado um modelo de regressão para dados de sobrevivência grupados na qual a abordagem é fundamentada em modelos de tempo discreto e em tabelas de vida. A estrutura de regressão representada por uma probabilidade é modelada adotando-se diferentes funções de ligação, tais como, logito, complemento log-log, log-log e probito. Em ambas as pesquisas, métodos de validação dos modelos estatísticos propostos são descritos e fundamentados na análise de sensibilidade. Para detectar observações influentes nos modelos propostos, foram utilizadas medidas de diagnóstico baseadas na deleção de casos, denominadas de influência global e medidas baseadas em pequenas perturbações nos dados ou no modelo proposto, denominada de influência local. Para verificar a qualidade de ajuste do modelo e detectar pontos discrepantes foi realizada uma análise de resíduos nos modelos propostos. Os resultados desenvolvidos foram aplicados a dois conjuntos de dados reais. / In this study, a regression model for interval-censored data were developed, using the Exponentiated- Weibull distribution, that has as main characteristic the hazard function which assumes different forms (unimodal, bathtub shape, increase, decrease). A good feature of that regression model is their use to discriminate models, that have as particular cases, the models of regression: Exponential, Weibull, Exponential-exponentiated, amongst others. Also a regression model were studied for grouped survival data in which the approach is based in models of discrete time and in life tables, the regression structure represented by a probability is modeled through the use of different link function, logit, complementary log-log, log-log or probit. In both studies, validation methods for the statistical models studied are described and based on the sensitivity analysis. To find influential observations in the studied models, diagnostic measures were used based on case deletion, denominated as global influence and measures based on small perturbations on the data or in the studied model, denominated as local influence. To verify the goodness of fitting of the model and to detect outliers it was performed residual analysis for the proposed models. The developed results were applied to two real data sets.

Dans le cadre de l’analyse de signaux mécaniques ou biomécaniques les outils d’aide à la décision reposent sur des hypothèses statistiques fortes: loi de probabilité normale, stationnarité des variables, variables centrées, variables indépendantes,…Or ces hypothèses sont parfois non vérifiées et engendrent des décisions erronées. Ce travail a pour objectif de proposer des méthodes qui font abstractions de certaines hypothèses et notamment de la stationnarité et de la gaussiannité des variables aléatoires. Dans cette thèse, nous avons revisité certaines méthodes de ré échantillonnages statistiques et de bootstrap et développé d’autres en incluant la cyclostationnarité des signaux. Ensuite, nous avons appliqué ces méthodes pour l’analyse de signaux biomécaniques provenant de coureurs expérimentés et d’une population de personnes âgées. Les résultats obtenus ont permis de mettre en évidence des changements significatifs dans le contenu fréquentiel du second ordre des signaux étudiés. Ces changements ont été des indicateurs très pertinents pour la description et la caractérisation de la fatigue d’un coureur professionnel, d’une part, et pour la compréhension du mécanisme complexe de la marche à pied simple et avec tâche cognitive chez les personnes âgées d’autre part / In mechanical and biomechanical signal analysis field, the decision support tools are based on strong statistical assumptions such as: normality, stationarity of variables, independence... However, these assumptions are very often unverified, consequently, wrong decisions could be taken. This work aims to propose new methods that make abstractions of such assumptions, including the stationarity and gaussianity of variables. In this thesis, we revisited some statistical resampling methods and developed new bootstrap approaches with including the cyclostationary nature of signals. Next, we applied these methods to the analysis of biomechanical signals from experienced runners and a population of elderly people. The obtained results allowed us to demonstrate significant changes in the second order frequency content of the signals under study. These changes were very relevant indicators for the description and characterization of the fatigue of a high level professional runner. Moreover, these changes helped us to understand the mechanism of normal walking and under a cognitive task condition (double task walking) in elderly

In this work, its presented a novel particle resampling method for free-surface fitting of liquids from particle-based fluid simulations. The proposed approach is simple and easy to implement, and only requires the positions of the particles to properly identify and refine regions with small-scale features. The method comprises three main stages: boundary detection, feature classification, and particle refinement. For each simulation frame, firstly the free-surface is captured through a boundary detection scheme as chosen by the user. Then, the boundary particles are classified and labeled according to the deformation and the stretching of the freesurface computed from the Principal Component Analysis (PCA) of the particle positions. Finally, particles placed at feature regions are refined according to their feature classification. In order to render the free-surface, its demonstrated how the traditional methods of free-surface fitting in Computer Graphics and Computational Physics literature can be benefited by the proposed resampling method. Moreover, the results shown in this work attest the effectiveness and robustness of the method when compared against state-of-the-art adaptive particle sampling techniques. / Neste trabalho é apresentado um novo método de re-amostragem para a superfície livre de liquidos de simulações de fluidos baseados em partícula. A abordagem proposta é simples e fácil de ser implementada, e requere apenas as posições das partículas para identificar e refinar devidamente regiões com características de baixa escala. O método consiste de três estágios principais: detecção de borda, classificação de caracteristicas e refinamento de partículas. Para cada quadro da simulação, primeiro a superfície livre é identificada a partir da detecção de borda, cujo método pode ser escolhido pelo usuário. Então, as partículas de borda são classificadas e recebem um valor de acordo com a deformação e alongamento da superfície livre calculada a partir da Análise de Componentes Principais (PCA) das posições das partículas. Por fim, partículas que pertencem à camadas finas são refinadas de acordo com suas classificações. Para renderizar a superfície livre, é demonstrado como métodos tradicionais de \"fitting\" de superfície em Computação Gráfica e Física computacional podem ser beneficiadas pelo método de reamostragem proposta. Por fim, os resultados mostrados neste trabalho certificam a efetividade e a robustez do método quando comparado com as técnicas adaptativas de amostragem de partículas do estado-da-arte.

We live in a dynamic world full of noises and interferences. The conventional acoustic echo cancellation (AEC) framework based on the least mean square (LMS) algorithm by itself lacks the ability to handle many secondary signals that interfere with the adaptive filtering process, e.g., local speech and background noise. In this dissertation, we build a foundation for what we refer to as the system approach to signal enhancement as we focus on the AEC problem. We first propose the residual echo enhancement (REE) technique that utilizes the error recovery nonlinearity (ERN) to "enhances" the filter estimation error prior to the filter adaptation. The single-channel AEC problem can be viewed as a special case of semi-blind source separation (SBSS) where one of the source signals is partially known, i.e., the far-end microphone signal that generates the near-end acoustic echo. SBSS optimized via independent component analysis (ICA) leads to the system combination of the LMS algorithm with the ERN that allows for continuous and stable adaptation even during double talk. Second, we extend the system perspective to the decorrelation problem for AEC, where we show that the REE procedure can be applied effectively in a multi-channel AEC (MCAEC) setting to indirectly assist the recovery of lost AEC performance due to inter-channel correlation, known generally as the "non-uniqueness" problem. We develop a novel, computationally efficient technique of frequency-domain resampling (FDR) that effectively alleviates the non-uniqueness problem directly while introducing minimal distortion to signal quality and statistics. We also apply the system approach to the multi-delay filter (MDF) that suffers from the inter-block correlation problem. Finally, we generalize the MCAEC problem in the SBSS framework and discuss many issues related to the implementation of an SBSS system. We propose a constrained batch-online implementation of SBSS that stabilizes the convergence behavior even in the worst case scenario of a single far-end talker along with the non-uniqueness condition on the far-end mixing system. The proposed techniques are developed from a pragmatic standpoint, motivated by real-world problems in acoustic and audio signal processing. Generalization of the orthogonality principle to the system level of an AEC problem allows us to relate AEC to source separation that seeks to maximize the independence, hence implicitly the orthogonality, not only between the error signal and the far-end signal, but rather, among all signals involved. The system approach, for which the REE paradigm is just one realization, enables the encompassing of many traditional signal enhancement techniques in analytically consistent yet practically effective manner for solving the enhancement problem in a very noisy and disruptive acoustic mixing environment.

This research develops efficient solution methods for linear systems with scalar and multi-level Toeplitz structure. Toeplitz systems are common in one-dimensional signal-processing applications, and typically correspond to temporal- or spatial-invariance in the underlying physical phenomenon. Over time, a number of algorithms have been developed to solve these systems economically by exploiting their structure. These developments began with the Levinson-Durbin recursion, a classical fast method for solving Toeplitz systems that has become a standard algorithm in signal processing. Over time, more advanced routines known as superfast algorithms were introduced that are capable of solving Toeplitz systems with even lower asymptotic complexity. For multi-dimensional signals, temporally- and spatially-invariant systems have linear-algebraic descriptions characterized by multi-level Toeplitz matrices, which exhibit Toeplitz structure on multiple levels. These matrices lack the same algebraic properties and structural simplicity of their scalar analogs. As a result, it has proven exceedingly difficult to extend the existing scalar Toeplitz algorithms for their treatment. This research presents algorithms to solve scalar and two-level Toeplitz systems through a constructive approach, using methods devised for specialized cases to build more general solution methods. These methods extend known scalar Toeplitz inversion results to more general scalar least-squares problems and to multi-level Toeplitz problems. The resulting algorithms have the potential to provide substantial computational gains for a large class of problems in signal processing, such as image deconvolution, non-uniform resampling, and the reconstruction of spatial volumes from non-uniform Fourier samples.

This thesis considers the filtering and prediction problems of nonlinear noisy econometric systems. As a filter/predictor, the standard tool Extended Kalman Filter and new approaches Discrete Quantization Filter and Sequential Importance Resampling Filter are used. The algorithms are compared by using Monte Carlo Simulation technique. The advantages of the new algorithms over Extended Kalman Filter are shown.

Introduction: In Positron Emission Tomography (PET), the use of resolution modelling (RM) in iterative image reconstruction enables the modelling of aspects of detection which result in mispositioning of measured data and the subsequent blurring of reconstructed images. RM reconstruction can result in significant improvements in spatial resolution, voxel variance and count rate bias and could be a software alternative to detection hardware that is able to achieve higher resolution. Such hardware typically consists of small scintillation crystals, small bore diameters and depth of interaction discrimination, such as for the High Resolution Research Tomograph (HRRT, Siemens), which used a double crystal layer phoswich detector system. However, RM implementation comes with penalties such as slower rates of convergence, potentially higher region of interest variance and Gibbs artefacts. Methods: Assessment of the benefits and drawbacks of RM was done in the first part of this thesis together with the measurement and modelling of spatially varying resolution kernels for different scanner configurations and PET isotopes for the HRRT. It is also unclear as to whether high resolution scanning offers significant advantages over clinical PET-CT scanners for applications in the head. Through direct comparison to our HRRT, we explore whether there are significant advantages of high resolution scanning for an application in the head over clinical PET-CT. For this comparison our Biograph TruePoint TrueV (Siemens) optimised for whole body imaging was used and a novel clinical study using both scanners was set where we scanned Neurofibromatosis 2 (NF2) patients with vestibular schwannomas (VS) using [18F]fluorodeoxyglucose (FDG) and [18F]fluorothymidine (FLT). The clinical objective was to assess if uptake within VS of FLT and FDG could be measured and whether this uptake was predictive of tumour growth. Finally an assessment of the feasibility and impact of reducing the original injected activities in our clinical study was performed using bootstrapping resampling. Conclusions: RM provides greater but additive improvements in image resolution compared to DOI on the HRRT. Isotope specific image based RM could be estimated from published positron range distributions and measurements using fluorine-18. With the clinical project, uptake of FDG and FLT within the VS lesions was observed, these uptake values were correlated to each other, and high uptake was predictive of tumour growth with little differences in predictive power between FLT and FDG. Although there were benefits of the HRRT for imaging small lesions, in our clinical application there was little difference between the two scanners to discriminate lesion growth. Using the PET-CT scanner data and knowledge of lesion location, doses could be reduced to 5-10% without any significant loss of ability to discriminate lesion growth.

Made available in DSpace on 2016-08-12T19:15:22Z (GMT). No. of bitstreams: 1 kallyo halyson santos moura ok.pdf: 596520 bytes, checksum: 9c71d52ee07aced37426fdeafa2cef96 (MD5) Previous issue date: 2008-04-30 / Conselho Nacional de Desenvolvimento Científico e Tecnológico / The sample size in experiments should be determined so that it can accurately estimate the parameters of interest and to save time, labor and resources. The objective of the present work was to determine the sample size in evaluation experiments of melon hybrids in Mossoró. An experiment was carried out in randomized blocks with three replications to evaluate twenty yellow melon hybrids. The experimental plot consisted of two rows of 5.0 m with twenty plants in the spacing of 2.0 x 0.5 m. The traits evaluated were pulp thickness and firmness, content of soluble solids and fruit mean weight. Data were taken plant to plant, identifying each one harvested fruits. The methods used were of bootstrap resampling, linear model in plateau, sampling ntensity and modified maximum curvature to determine the sample size. The sample size recommended to determining for pulp thickness, fruit mean weight, content of soluble solidsand pulp firmness is 13, 13, 9 and 8 fruits, respectively. The Minimum sample size determined by modified maximum curvature method to pulp thickness, fruit mean weight, content of soluble solids and pulp firmness were 6,32; 7,35; 5,32and 3,25, respectively / O tamanho da amostra em experimentos deve ser determinado para que se possa estimar com precisão adequada os parâmetros de interesse e economizar tempo, mão-de-obra e recursos. O objetivo do presente trabalho foi determinar o tamanho da amostra para avaliação de híbridos de melão amarelo. Foi conduzido um experimento em blocos casualizados com três repetições para avaliar vinte híbridos de melão. A parcela utilizada foi constituída por duas linhas de 5,0 m com vinte plantas no espaçamento de 2,0 x 0,5 m. As características avaliadas foram as seguintes: espessura da polpa, firmeza da polpa, teor de sólidos solúveis e peso médio do fruto. Os dados foram tomados planta a planta, identificando cada um dos frutos colhidos. Foram utilizados os métodos de reamostragem bootstrap, modelo linear segmentado com platô, intensidade de amostragem e máxima curvatura para determinar o tamanho da amostra. O tamanho de amostra recomendado para determinação da espessura da polpa, peso médio do fruto, teor de sólidos solúveis e firmeza da polpa é de 13, 13, 9 e 8 frutos, respectivamente. Os tamanhos mínimos da amostra, obtidos pelo método da máxima curvatura modificada, foram 6,32; 7,35; 5,32 e 3,25 para espessura da polpa, peso médio do fruto, teor de sólidos solúveis e firmeza da polpa, respectivamente

Made available in DSpace on 2016-08-12T19:18:43Z (GMT). No. of bitstreams: 1 kallyo halyson santos moura ok.pdf: 596520 bytes, checksum: 9c71d52ee07aced37426fdeafa2cef96 (MD5) Previous issue date: 2008-04-30 / Conselho Nacional de Desenvolvimento Científico e Tecnológico / The sample size in experiments should be determined so that it can accurately estimate the parameters of interest and to save time, labor and resources. The objective of the present work was to determine the sample size in evaluation experiments of melon hybrids in Mossoró. An experiment was carried out in randomized blocks with three replications to evaluate twenty yellow melon hybrids. The experimental plot consisted of two rows of 5.0 m with twenty plants in the spacing of 2.0 x 0.5 m. The traits evaluated were pulp thickness and firmness, content of soluble solids and fruit mean weight. Data were taken plant to plant, identifying each one harvested fruits. The methods used were of bootstrap resampling, linear model in plateau, sampling ntensity and modified maximum curvature to determine the sample size. The sample size recommended to determining for pulp thickness, fruit mean weight, content of soluble solidsand pulp firmness is 13, 13, 9 and 8 fruits, respectively. The Minimum sample size determined by modified maximum curvature method to pulp thickness, fruit mean weight, content of soluble solids and pulp firmness were 6,32; 7,35; 5,32and 3,25, respectively / O tamanho da amostra em experimentos deve ser determinado para que se possa estimar com precisão adequada os parâmetros de interesse e economizar tempo, mão-de-obra e recursos. O objetivo do presente trabalho foi determinar o tamanho da amostra para avaliação de híbridos de melão amarelo. Foi conduzido um experimento em blocos casualizados com três repetições para avaliar vinte híbridos de melão. A parcela utilizada foi constituída por duas linhas de 5,0 m com vinte plantas no espaçamento de 2,0 x 0,5 m. As características avaliadas foram as seguintes: espessura da polpa, firmeza da polpa, teor de sólidos solúveis e peso médio do fruto. Os dados foram tomados planta a planta, identificando cada um dos frutos colhidos. Foram utilizados os métodos de reamostragem bootstrap, modelo linear segmentado com platô, intensidade de amostragem e máxima curvatura para determinar o tamanho da amostra. O tamanho de amostra recomendado para determinação da espessura da polpa, peso médio do fruto, teor de sólidos solúveis e firmeza da polpa é de 13, 13, 9 e 8 frutos, respectivamente. Os tamanhos mínimos da amostra, obtidos pelo método da máxima curvatura modificada, foram 6,32; 7,35; 5,32 e 3,25 para espessura da polpa, peso médio do fruto, teor de sólidos solúveis e firmeza da polpa, respectivamente

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Calendula is an important plant cultivated in southern Brazil with medicinal properties, therapeutic and for increasing farmers' income. The goals of the study were to estimate the morphological characters like flower height (cm), plant height (cm), number of leaves per plant and diameter of the flower head (cm) of the main stem, also the productive characters as dry mass (g plant-1) and seed mass (g plant-1); ascertain the viability of seed production for the species, the duration of growth periods and development in growing seasons; determine sample sizes for average estimation of the characters mentioned above, with the exception of the seed mass; check the variability of the sample sizes between the characters and evaluation periods and; evaluate the genotoxic and anti-proliferative effects of calendula infusions on the in vivo test system of Allium cepa. It was conducted a field experiment in a randomized block design with three factors which are: growing seasons (August/14, October/14 and April/15), cultivars (Bonina Dobrada Sortida and Bon Bon Yellow) and cropping systems (direct seeding and transplanting seedlings), in a total of twelve treatments. The results indicate greater cycle, flowering period and growth to the period April/15. Cultivar Sortida was superior than Yellow for any combination of seasons and cropping systems for flower height character ranging between 26.84 cm to 39.76 cm. Plants originated by direct seeding demonstrated superiority for flower height for season April/15, regardless of the cultivar. This system was also higher for characters like plant height, leaf number, and dry mass and mass of seeds at the time April/15. For plant height, leaf number and dry weight of the plant growing seasons in August/14 October/14 and April/15 the variety Sortida had the highest averages. In relation to growing seasons, the flower head diameter character had the highest average in the period April/15 of 6.82 cm. Among the cultivars, Sortida differed from variety Yellow by showing 7.03 cm average. The germination percentages of the seeds harvested in three seasons under study were very low for use by the producer. Sample sizes were very different between the characters, and for plant height and number of leaves values were similar in the evaluation performed soon after transplanting and flowering plants. There were necessary smaller sample sizes for morphological characteristics in relation to dry mass production. Evaluation of 190 plants is enough to estimate the mean amplitude of the parameters evaluated for a 95% confidence interval with a maximum error of 20% of the mean estimate. Infusions derived from cut flowers of the season in August/14 showed antiproliferative effect, reducing the mitotic index compared to controls for all infusions, it is not possible to determine whether there are genotoxic effects. / A calêndula é uma espécie cultivada no sul do Brasil com importância medicinal, terapêutica e no incremento de renda dos agricultores. Os objetivos do trabalho foram: estimar os caracteres morfológicos altura do capítulo (cm), altura de planta (cm), número de folhas por planta e diâmetro do capítulo (cm) da haste principal, além dos caracteres produtivos massa seca (g planta-1) e massa de sementes (g planta-1); verificar a viabilidade da produção de sementes da espécie; averiguar a duração dos períodos de crescimento e desenvolvimento em épocas de cultivo; determinar tamanhos de amostra para estimação da média dos caracteres mencionados acima, com exceção da massa de sementes; investigar a variabilidade dos tamanhos de amostra entre os caracteres e épocas de avaliação e; avaliar o efeito genotóxico e antiproliferativo de infusões de calêndula sobre o sistema teste in vivo de Allium cepa. Conduziu-se um experimento em campo no delineamento blocos ao acaso com três fatores: épocas de cultivo (agosto/14, outubro/14 e abril/15), cultivares (Bonina Dobrada Sortida e Bon Bon Yellow) e sistemas de cultivo (semeadura direta e transplante de mudas), totalizando doze tratamentos. Os resultados indicaram maior ciclo, período de floração e crescimento na época abril/15. A cultivar Bonina Dobrada Sortida foi superior em qualquer combinação das épocas e sistemas de cultivo para o caractere altura de flor variando entre 26,84 cm e 39,76 cm. Plantas oriundas da semeadura direta mostraram superioridade para altura de flor na época abril/15, independente da cultivar. Esse sistema também foi superior para os caracteres altura de planta, número de folhas, massa seca e massa de sementes na época abril/15. Para altura de planta, número de folhas e massa seca nas épocas de cultivo agosto/14, outubro/14 e abril/15 a cultivar Bonina Dobrada Sortida apresentou as maiores médias. Em relação às épocas de cultivo, o caractere diâmetro do capítulo apresentou a maior média na época abril/15 com 6,82 cm. Dentre as cultivares a Bonina Dobrada Sortida diferiu da cultivar Bom Bon Yellow, apresentando 7,03 cm de média. As porcentagens de germinação das sementes colhidas das três épocas em estudo foram muito baixas para utilização pelo produtor. Os tamanhos de amostra foram bastante distintos entre os caracteres, sendo que para altura de plantas e número de folhas os valores foram semelhantes na avaliação realizada logo após o transplante e no florescimento das plantas. Verificaram-se tamanhos de amostra menores para caracteres morfológicos em relação ao produtivo massa seca. A avaliação de 190 plantas é suficiente para estimar a média dos caracteres avaliados para amplitude do intervalo de confiança de 95% com erro máximo de 20% da estimativa da média. As infusões oriundas de flores colhidas da época agosto/14 apresentaram efeito antiproliferativo, com redução do índice mitótico em relação aos controles para todas as infusões, não sendo possível determinar se existem efeitos genotóxicos.