Esta dissertação compreende um estudo da eficiência de estimadores dos parâmetros no modelo funcional com erro nas variáveis, com repetições para contornar o problema de falta de identificação. Nela, discute-se os procedimentos baseados nos métodos de máxima verossimilhança e escore corrigido. As estimativas obtidas pelos dois métodos levam a resultados similares. / This work is concerned with a study on the efficiency of parameter estimates in the functional linear relashionship with constant variances. Where the lack of identification is resolved of by considering replications. Estimation is dealt with by using maximum likelihood and the corrected score approach. Comparisons between the approaches are illustrated by using simulated data.

This thesis describes the use of scattering functions to characterize time-varying multipath radio channels. Channel Impulse responses were measured at Edwards Air Force Base (EAFB) and scattering functions generated from the impulse response data. From the scattering functions we compute the corresponding Doppler power spectrum and multipath intensity profile. These functions completely characterize the signal delay and the time varying nature of the channel in question and are used by systems engineers to design reliable communications links. We observe from our results that flight paths with ample reflectors exhibit significant multipath events. We also examine the bit error rate (BER) performance of a reduced-complexity equalizer for a truncated version of the pulse amplitude modulation (PAM) representation of SOQPSK-TG in a multipath channel. Since this reduced-complexity equalizer is based on the maximum likelihood (ML) principle, we expect it to perform optimally than any of the filter-based equalizers used in estimating received SOQPSK-TG symbols. As such we present a comparison between this ML detector and a minimum mean square error (MMSE) equalizer for the same example channel. The example channel used was motivated by the statistical channel characterizations described in thisthesis. Our analysis shows that the ML equalizer outperforms the MMSE equalizer in estimating received SOQPSK-TG symbols.

In the event of a natural or anthropogenic disturbance, environmental resource managers require a reliable tool to quickly assess the spatial extent of potential damage to the seagrass resource. The temporal availability of the Landsat 5 Thematic Mapper (TM) imagery, 16-20 days, provides a suitable option to detect and assess damage to the seagrass resource. In this study, remote sensing Landsat 5 TM imagery is used to map the spatial extent of the seagrass resource. Various classification techniques are applied to delineate the seagrass beds in Clearwater Harbor and St. Joseph Sound, FL. This study aims to determine the most appropriate seagrass habitat mapping technique by evaluating the accuracy and validity of the resultant classification maps. Field survey data and high resolution aerial photography are available to use as ground truth information. Seagrass habitat in the study area consists of seagrass species and rhizophytic algae; thus, the species assemblage is categorized as submerged aquatic vegetation (SAV). Two supervised classification techniques, Maximum Likelihood and Mahalanobis Distance, are applied to extract the thematic features from the Landsat imagery. The Mahalanobis Distance classification (MDC) method achieves the highest overall accuracy (86%) and validation accuracy (68%) for the delineation of the presence/absence of SAV. The Maximum Likelihood classification (MLC) method achieves the highest overall accuracy (74%) and validation accuracy (70%) for the delineation of the estimated coverage of SAV for the classes of continuous and patchy seagrass habitat. The soft classification techniques, linear spectral unmixing (LSU) and artificial neural network (ANN), did not produce reasonable results for this particular study. The comparison of the MDC and MLC to the current Seagrass Aerial Photointerpretation (AP) project indicates that the classification of SAV from Landsat 5 TM imagery provides a map product with similar accuracy to the AP maps. These results support the application of remote sensing thematic feature extraction methods to analyze the spatial extent of the seagrass resource. While the remote sensing thematic feature extraction methods from Landsat 5 TM imagery are deemed adequate, the use of hyperspectral imagery and better spectral libraries may improve the identification and mapping accuracy of the seagrass resource.

Although the captive reproduction of non-avian reptiles has increased steadily since the 1970’s, a dearth of information exists on successful management practices for large captive populations of these species. The data reported here come from a captive population of ball pythons (Python regius) maintained by a commercial breeding company, The Snake Keeper, Inc. (Spanish Fork, UT). Reproductive data are available for 6,480 eggs from 937 ball python clutches. The data presented suggest that proper management practices should include the use of palpation and/or ultrasound to ensure breeding occurs during the proper time of the female reproductive cycle, and that maintenance of proper humidity during the incubation of eggs is vitally important. Ball python reproduction traits (clutch size, clutch mass, relative clutch mass, egg mass, hatch rate, egg length, egg width, hatchling mass, healthy offspring per clutch, week laid, and days of incubation) were recorded for the clutches laid during this study. For the 937 clutches, the identity of the dam and sire were known for 862 (92%) and 777 (83%) of the clutches, respectively. A multivariate model that included nine of the 11 traits listed above was compiled. Heritability and genetic and phenotypic correlations were calculated from the multivariate analysis. The trait that showed the most promise for use in artificial selection to increase reproduction rates was clutch size due to considerable genetic variation, high heritability, and favorable genetic correlations with other reproduction traits. Although large datasets have been published for twinning in avian species, relatively few are available for non-avian reptiles. Reported here are 14 sets of twins produced from 6,480 eggs from 937 ball python clutches. The survival rate for twins during the first 3 months of life in our study was 97%. Interestingly, 11 of the sets of twins were identical in sex and phenotype, and additional genetic data suggested the rate of monozygotic twinning within this captive population of ball pythons was higher than that of dizygotic twinning. Further, using microsatellite analysis we were able to generate data that shows three sets of python twins were genetically identical.

Em análise de dados que apresentam certo grau de assimetria, curtose ou bimodalidade, a suposição de normalidade não é válida, sendo necessários modelos que capturem estas características dos dados. Neste contexto, uma nova classe de distribuições bimodais assimétricas gerada por um mecanismo de mistura é proposta neste trabalho. Algumas propriedades para o caso particular que inclui a distribuição normal como família base desta classe são estudadas e apresentadas, tal caso resulta no chamado Modelo Mistura de Potência Normal (MPN). Dois algoritmos de simulação são desenvolvidos com a finalidade de obter variáveis aleatórias com esta distribuição. A abordagem frequentista é empregada para a inferência dos parâmetros do modelo proposto. São realizados estudos de simulação com o objetivo de avaliar o comportamento das estimativas de máxima verossimilhança dos parâmetros. Adicionalmente, um modelo de regressão para dados bimodais é proposto, utilizando a distribuição MPN como variável resposta nos modelos Generalizados Aditivos para Posição, Escala e Forma, cuja sigla em inglês é GAMLSS. Para este modelo de regressão estudos de simulação também são realizados. Em ambos os casos estudados, o modelo proposto é ilustrado utilizando um conjunto de dados reais referente à pontuação de jogadores na Super Liga Brasileira de Voleibol Masculino 2014/2015. Com relação a este conjunto de dados, o modelo MPN apresenta melhor ajuste quando comparado à modelos já existentes na literatura para dados bimodais. / In analysis of data that present a certain degree of asymmetry, kurtosis or bimodality, the assumption of normality is not valid and models that capture these characteristics of the data are required. In this context, a new class of bimodal asymmetric distributions generated by a mixture mechanism is proposed. Some properties for the particular case that includes the normal distribution as the base family of this class are studied and presented, such case results in the so-called Power Normal Mixture Model. Two simulation algorithms are developed with the purpose of obtaining random variables with this new distribution. The frequentist approach is used to the inference of the model parameters. Simulation studies are carried out with the aim of assessing the behavior of the maximum likelihood estimates of the parameters. In addition, the power normal mixture distribution is introduced as the response variable for the Generalized Additives Models for Location, Scale and Shape (GAMLSS). For this regression model, simulation studies are also performed. In both cases studied, the proposed model is illustrated using a data set on players\' scores in the Male Brazilian Volleyball Superliga 2014/2015. With respect to this dataset, the power normal mixture model presents better fit when compared to models already existing in the literature to bimodal data.

Hyperspectral remote sensing image data, such as that recorded by AVIRIS with 224 spectral bands, provides rich information on ground cover types. However, it presents new problems in machine assisted interpretation, mainly in long processing times and the difficulties of class training due to the low ratio of number of training samples to the number of bands. This thesis investigates feasible and efficient feature reduction and image classification techniques which are appropriate for hyperspectral image data. The study is reported in three parts. The first concerns a deterministic approach for hyperspectral data interpretation. Multigroup and multiple threshold spectral coding procedures, and associated techniques for spectral matching and classification, are proposed and tested. By coding on subgroups of bands using one or three thresholds, spectral searching and matching becomes simple, fast and free of the need for radiometric correction. Modifications of existing statistical techniques are proposed in the second part of the investigation A block-based maximum likelihood classification technique is developed. Several subgroups are formed from the complete set of spectral bands in the data, based on the properties of global correlation among the bands. Subgroups which are poorly correlated with each other are treated independently using conventional maximum likelihood classification. Experimental results demonstrate that, when using appropriate subgroup sizes, the new method provides a compromise among classification accuracy, processing time and available training pixels. Furthermore, a segmented, and possibly multi-layer, principal components transformation is proposed as a possible feature reduction technique prior to classification, and for effective colour display. The transformation is performed efficiently on each of the highly correlated subgroups of bands independently. Selected features from each transformed subgroup can be then transformed again to achieve a satisfactory data reduction ratio and to generate the three most significant components for colour display. Classification accuracy is improved and high quality colour image display is achieved in experiments using two AVIRIS data sets.

octor of Philosophy(PhD) / Molecular Evolution is the key to explain the divergence of species and the origin of life on earth. The main task in the study of molecular evolution is the reconstruction of evolutionary trees from sequences data of the current species. This thesis introduces a novel algorithm for inferring evolutionary trees from genetic data using quartet-based approach. The new method recursively merges sub-trees based on a global statistical provided by the global quartet weight matrix. The quarte weights can be computed using several methods. Since the quartet weights computation is the most expensive procedure in this approach, the new method enables the parallel inference of large evolutionary trees. Several techniques developed to deal with quartets inaccuracies. In addition, the new method we developed is flexible in such a way that can combine morphological and molecular phylogenetic analyses to yield more accurate trees. Also, we introduce the concept of critical point where more than one possible merges are possible for the same sub-tree. The critical point concept can provide information about the relationships between species in more details and show how close they are. This enables us to detect other reasonable trees. We evaluated the algorithm on both synthetic and real data sets. Experimental results showed that the new method achieved significantly better accuracy in comparison with existing methods.

Abstract This thesis considers some linear and nonlinear time series models. In the linear case, the analysis of a large number of short time series generated by a first order autoregressive type model is considered. The conditional and exact maximum likelihood procedures are developed to estimate parameters. Simulation results are presented and compare the bias and the mean square errors of the parameter estimates. In Chapter 3, five important nonlinear models are considered and their time series properties are discussed. The estimating function approach for nonlinear models is developed in detail in Chapter 4 and examples are added to illustrate the theory. A simulation study is carried out to examine the finite sample behavior of these proposed estimates based on the estimating functions.

Real-world learning tasks often involve high-dimensional data sets with complex patterns of missing features. In this paper we review the problem of learning from incomplete data from two statistical perspectives---the likelihood-based and the Bayesian. The goal is two-fold: to place current neural network approaches to missing data within a statistical framework, and to describe a set of algorithms, derived from the likelihood-based framework, that handle clustering, classification, and function approximation from incomplete data in a principled and efficient manner. These algorithms are based on mixture modeling and make two distinct appeals to the Expectation-Maximization (EM) principle (Dempster, Laird, and Rubin 1977)---both for the estimation of mixture components and for coping with the missing data.

The dependence between survival times and covariates is described e.g. by proportional hazard models. We consider partly parametric Cox models and discuss here the estimation of interesting parameters. We represent the ma- ximum likelihood approach and extend the results of Huang (1999) from linear to nonlinear parameters. Then we investigate the least squares esti- mation and formulate conditions for the a.s. boundedness and consistency of these estimators.