Indiana University-Purdue University Indianapolis (IUPUI) / Randomized studies are designed to estimate the average treatment effect (ATE) of an intervention. Individuals may derive quantitatively, or even qualitatively, different effects from the ATE, which is called the heterogeneity of treatment effect. It is important to detect the existence of heterogeneity in the treatment responses, and identify the different sub-populations. Two corresponding statistical methods will be discussed in this talk: a hypothesis testing procedure and a mixture-model based approach. The hypothesis testing procedure was constructed to test for the existence of a treatment effect in sub-populations. The test is nonparametric, and can be applied to all types of outcome measures. A key innovation of this test is to build stochastic search into the test statistic to detect signals that may not be linearly related to the multiple covariates. Simulations were performed to compare the proposed test with existing methods. Power calculation strategy was also developed for the proposed test at the design stage. The mixture-model based approach was developed to identify and study the sub-populations with different treatment effects from an intervention. A latent binary variable was used to indicate whether or not a subject was in a sub-population with average treatment benefit. The mixture-model combines a logistic formulation of the latent variable with proportional hazards models. The parameters in the mixture-model were estimated by the EM algorithm. The properties of the estimators were then studied by the simulations. Finally, all above methods were applied to a real randomized study in a low ejection fraction population that compared the Implantable Cardioverter Defibrillator (ICD) with conventional medical therapy in reducing total mortality.

Orientador: Celso Pascoli Bottura / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e Computação / Made available in DSpace on 2018-08-31T09:18:30Z (GMT). No. of bitstreams: 1 Serra_GinalberLuizdeOliveira_D.pdf: 2165582 bytes, checksum: a1dad46bc4d817f8d4e6457f60ae9599 (MD5) Previous issue date: 2005 / Resumo: Esta tese apresenta propostas de metodologias para identificação e controle inteligentes. Uma metodologia para identificação de sistemas dinâmicos não-lineares no tempo discreto, baseada tio método de variável instrumental e no modelo nebuloso Takagi-Sugeno, é apresentada. Nesta metodologia, a qual é uma extensão do método de variável instrumental tradicional, as variáveis instrumentais escolhidas, estatisticamente independentes do ruído, são mapeadas em conjuntos nebulosos, particionando o espaço de entrada em sub-regiões, para estimação não-polarizada dos parâmetros do conseqüente dos modelos nebulosos TS em ambiente ruidoso. Um esquema de controle adaptativo gain scheduling baseado em redes neurais, sistemas nebulosos e algoritmos genéticos para sistemas dinâmicos não-lineares no tempo discreto também é apresentado. 0 controlador nebuloso é desenvolvido e projetado com o usa de um algoritmo genético para satisfazer, simultaneamente, múltiplos objetivos. Com o esquema de aprendizagem supervisionada, os parâmetros do controlador nebuloso são usados para projetar um gain scheduler neural para ajuste on-line do controlador nebuloso em alguns pontos de operação do sistema dinâmico / Abstract: This thesis presents proposals of methodologies for intelligent identification and control. A methodology tor nonlinear dynamic discrete time systems identification, based on the instrumental variable method and Takagi-Sugeno fuzzy model, is presented. In this methodology, which is an extension of the standard instrumental variable method, the chosen instrumental variables, estatistically independent of the noise, are mapped into fuzzy sets, partitioning the input space in subregions, for unbiased estimation of Takagi-Sugeno fuzzy model consequent parameters in a noisy environment. A gain scheduling adaptive control design based on neural network, fuzzy systems and genetic algorithms for nonlinear dynamic discrete time systems is also presented. The fuzzy controller is developed and designed by a genetic algorithm to satisfy, simultaneously, multiple objectives. "With the supervised learning scheme, the fuzzy controller parameters are used to design the gain neural scheduler to tune on-line the fuzzy controller in some operation points of the dynamic system / Doutorado / Automação / Doutor em Engenharia Elétrica