[pt] Nesta dissertação é desenvolvido um modelo para previsão de curto prazo da demanda mensal de gasolina automotiva no Brasil. A metodologia usada consiste em, a partir de uma análise exploratória dos dados, procurar construir um modelo usando uma estratégia bottom-up, ou seja, parte-se de um modelo simples e processa-se seu refinamento até encontrar um modelo apropriado que mais se adequa à realidade. Partiu-se então de um modelo autoprojetivo indo até uma formulação de Redes Neurais passando por um modelo de regressão dinâmica. Os modelos são então comparados segundo alguns critérios, basicamente no que tange à sua eficiência preditiva. Conclui-se ao final sobre a eficiência de se conjugar modelos estatísticos clássicos (como Box & Jenkins e Regressão Dinâmica) com as chamadas Redes Neurais que, por sua vez, propiciaram resultados muito bons em relação à otimização das previsões. Isto é altamente desejável na modelagem de séries temporais e, em particular, neste trabalho, na previsão de curto prazo de gasolina automotiva. / [en] In this dissertation a short term model to forecast automotive gasoline demand in Brazil is proposed. From the methodology point of view, data is analyzed and a model using a bottom-up strategy is developed. In other words, a simple model is improved step by step until a proper model that sits well the reality is found. Departuring from a univariate model it ends up in a neural network formulation, passing through dynamic regression models. The models obtained in this scheme are compared according to some criterion, mainly forecast accuracy. We conclude, that the efficiency of putting together classical statistics models (such as Box & Jenkins and dynamic regression) and neural networks improve the forecasting results. This results is highly desirable in modeling time series and, particularly, to the short term forecast of automotive gasoline, object of this dissertation.

This research investigates route choice behavior in networks with risky travel times and real-time information. A stated preference survey is conducted in which subjects use a PC-based interactive maps to choose routes link-by-link in various scenarios. The scenarios include two types of maps: the first presenting a choice between one stochastic route and one deterministic route, and the second with real-time information and an available detour. The first type measures the basic risk attitude of the subject. The second type allows for strategic planning, and measures the effect of this opportunity on subjects' choice behavior. Results from each subject are analyzed to determine whether subjects planned strategically for the en route information or simply selected fixed paths from origin to destination. The full data set is used to estimate route choice models that account for both risk attitude and strategic thinking. Estimation results are used to assess whether models that incorporate strategic behavior more accurately reflect route choice than do simpler path-based models.

No description available.

No description available.

<p>The bending moduli and line tension of bilayer membranes self-assembled from diblock copolymers was calculated using the self-consistent field theory. The limitation of the linear elasticity theory (Helfrich model) was evaluated by calculating fourth- order curvature moduli in high curvature systems. It was found that in highly curved membranes, the fourth-order contributions to the bending energy becomes comparable to the low-order terms. The line tension (γL) of membrane pores was also investigated for mixtures of structurally different diblock copolymers. The line ten- sion was found to depend sensitively on the diblock chain topology. Addition of short hydrophobic copolymers was found to reduce the line tensions to negative values, showing that lipid mixtures may be used as pore stabilizers.</p> / Master of Science (MSc)

<p>As energy costs and energy used by server farms increase, so does the desire to implement energy-aware policies. Although under some cost functions, optimal policies for single as well as multiple server systems are known, large gaps in theoretical knowledge are present in the field. Specifically, there exists many widely used and non-trivial cost functions, where the corresponding optimal policy remains unknown. This work presents and leverages a model which allows for the exact analysis of these optimal policies with considerable generality, for on/off single server systems under a broad range of cost functions that are based on expected response time, energy usage and switching costs. Furthermore, from the results derived in the analysis, several applications and implications are presented and discussed. This includes the determination of routing probabilities to show a range of non-trivial optimal routing probabilities and server configurations when energy concerns are a factor.</p> / Master of Applied Science (MASc)

Cancer is a leading cause of death globally, and early detection is crucial for better outcomes. This research aims to improve Region Of Interest (ROI) segmentation and feature extraction in medical image analysis using Radiomics techniques with 3D Slicer, Pyradiomics, and Python. Dimension reduction methods, including PCA, K-means, t-SNE, ISOMAP, and Hierarchical Clustering, were applied to highdimensional features to enhance interpretability and efficiency. The study assessed the ability of the reduced feature set to predict T-staging, an essential component of the TNM system for cancer diagnosis. Multinomial logistic regression models were developed and evaluated using MSE, AIC, BIC, and Deviance Test. The dataset consisted of CT and PET-CT DICOM images from 131 lung cancer patients. Results showed that PCA identified 14 features, Hierarchical Clustering 17, t-SNE 58, and ISOMAP 40, with texture-based features being the most critical. This study highlights the potential of integrating Radiomics and unsupervised learning techniques to enhance cancer prediction from medical images.

Multi-state models have been widely used to analyze longitudinal event history data obtained in medical studies. The tools and methods developed recently in this area require the complete observed datasets. While, in many applications measurements on certain components of the covariate vector are missing on some study subjects. In this dissertation, several likelihood-based methodologies were proposed to deal with datasets with different types of missing covariates efficiently when applying multi-state models. Firstly, a maximum observed data likelihood method was proposed when the data has a univariate missing pattern and the missing covariate is a categorical variable. The construction of the observed data likelihood function is based on the model of a joint distribution of the response longitudinal event history data and the discrete covariate with missing values. Secondly, we proposed a maximum simulated likelihood method to deal with the missing continuous covariate when applying multi-state models. The observed data likelihood function was approximated by using the Monte Carlo simulation method. At last, an EM algorithm was used to deal with multiple missing covariates when estimating the parameters of multi-state model. The EM algorithm would be able to handle multiple missing discrete covariates in general missing pattern efficiently. All the proposed methods are justified by simulation studies and applications to the datasets from the SMART project, a consortium of 11 different high-quality longitudinal studies of aging and cognition.

Combining effect sizes from individual studies using random-effects models are commonly applied in high-dimensional gene expression data. However, unknown study heterogeneity can arise from inconsistency of sample qualities and experimental conditions. High heterogeneity of effect sizes can reduce statistical power of the models. We proposed two new methods for random effects estimation and measurements for model variation and strength of the study heterogeneity. We then developed a statistical technique to test for significance of random effects and identify heterogeneous genes. We also proposed another meta-analytic approach that incorporates informative weights in the random effects meta-analysis models. We compared the proposed methods with the standard and existing meta-analytic techniques in the classical and Bayesian frameworks. We demonstrate our results through a series of simulations and application in gene expression neurodegenerative diseases.

High-dimensional data are becoming increasingly available as data collection technology advances. Over the last decade, significant developments have been taking place in high-dimensional data analysis, driven primarily by a wide range of applications in many fields such as genomics, signal processing, and environmental studies. Statistical techniques such as dimension reduction and variable selection play important roles in high dimensional data analysis. Sufficient dimension reduction provides a way to find the reduced space of the original space without a parametric model. This method has been widely applied in many scientific fields such as genetics, brain imaging analysis, econometrics, environmental sciences, etc. in recent years. In this dissertation, we worked on three projects. The first one combines local modal regression and Minimum Average Variance Estimation (MAVE) to introduce a robust dimension reduction approach. In addition to being robust to outliers or heavy-tailed distribution, our proposed method has the same convergence rate as the original MAVE. Furthermore, we combine local modal base MAVE with a $L_1$ penalty to select informative covariates in a regression setting. This new approach can exhaustively estimate directions in the regression mean function and select informative covariates simultaneously, while being robust to the existence of possible outliers in the dependent variable. The second project develops sparse adaptive MAVE (saMAVE). SaMAVE has advantages over adaptive LASSO because it extends adaptive LASSO to multi-dimensional and nonlinear settings, without any model assumption, and has advantages over sparse inverse dimension reduction methods in that it does not require any particular probability distribution on \textbf{X}. In addition, saMAVE can exhaustively estimate the dimensions in the conditional mean function. The third project extends the envelope method to multivariate spatial data. The envelope technique is a new version of the classical multivariate linear model. The estimator from envelope asymptotically has less variation compare to the Maximum Likelihood Estimator (MLE). The current envelope methodology is for independent observations. While the assumption of independence is convenient, this does not address the additional complication associated with a spatial correlation. This work extends the idea of the envelope method to cases where independence is an unreasonable assumption, specifically multivariate data from spatially correlated process. This novel approach provides estimates for the parameters of interest with smaller variance compared to maximum likelihood estimator while still being able to capture the spatial structure in the data.