We present a theoretical approach to the statistical analysis of the dependence of the gap time length between consecutive recurrent events, on a set of explanatory random variables and in the presence of right censoring. The dependence is expressed through regression-like and overdispersion parameters, estimated via estimating functions and equations. The mean and variance of the length of each gap time, conditioned on the observed history of prior events and other covariates, are known functions of parameters and covariates, and are part of the estimating functions. Under certain conditions on censoring, we construct normalized estimating functions that are asymptotically unbiased and contain only observed data. We then use modern mathematical techniques to prove the existence, consistency and asymptotic normality of a sequence of estimators of the parameters. Simulations support our theoretical results.

Esta tese apresenta DCE, ou Execução Dinâmica Condicional, como uma alternativa para reduzir o custo da previsão incorreta de desvios. A idéia básica do modelo apresentado é buscar e executar todos os caminhos de desvios que obedecem à certas restrições no que diz respeito a complexidade e tamanho. Como resultado, tem-se um número menor de desvios sendo previstos e consequentemente um número menor de desvios previstos incorretamente. DCE busca todos os caminhos dos desvios selecionados evitando quebras no fluxo de busca quando estes desvios são buscados. Os caminhos buscados dos desvios selecionados são então executados mas somente o caminho correto é completado. Nesta tese nós propomos uma arquitetura para executar múltiplos caminhos dos desvios selecionados. A seleção dos desvios ocorre baseada no tamanho do desvio e em outras condições. A seleção de desvios simples e complexos permite a predicação dinâmica destes desvios sem a necessidade da existência de um conjunto específico de instruções nem otimizações especiais por parte do compilador. Além disso, é proposta também uma técnica para reduzir a sobrecarga gerada pela execução dos múltiplos caminhos dos desvios selecionados. O desempenho alcançado atinge níveis de até 12% quando um previsor de desvios Local é usado no DCE e um previsor Global é usado na máquina de referência. Quando ambas as máquinas empregam previsão Local, há um aumento de desempenho da ordem de 3-3.5%. / This thesis presents DCE, or Dynamic Conditional Execution, as an alternative to reduce the cost of mispredicted branches. The basic idea is to fetch all paths produced by a branch that obey certain restrictions regarding complexity and size. As a result, a smaller number of predictions is performed, and therefore, a lesser number of branches are mispredicted. DCE fetches through selected branches avoiding disruptions in the fetch flow when these branches are fetched. Both paths of selected branches are executed but only the correct path commits. In this thesis we propose an architecture to execute multiple paths of selected branches. Branches are selected based on the size and other conditions. Simple and complex branches can be dynamically predicated without requiring a special instruction set nor special compiler optimizations. Furthermore, a technique to reduce part of the overhead generated by the execution of multiple paths is proposed. The performance achieved reaches levels of up to 12% when comparing a Local predictor used in DCE against a Global predictor used in the reference machine. When both machines use a Local predictor, the speedup is increased by an average of 3-3.5%.

A Execução Condicional Dinâmica (DCE) é uma alternativa para redução dos custos relacionados a desvios previstos incorretamente. A idéia básica é buscar todos os fluxos produzidos por um desvio que obedecem algumas restrições relativas à complexidade e tamanho. Como conseqüência, um número menor de previsões é executado, e assim, um número mais baixo de desvios é incorretamente previsto. Contudo, tal como outras soluções multi-fluxo, o DCE requer uma estrutura de controle mais complexa. Na arquitetura DCE, é observado que várias réplicas da mesma instrução são despachadas para as unidades funcionais, bloqueando recursos que poderiam ser utilizados por outras instruções. Essas réplicas são geradas após o ponto de convergência dos diversos fluxos em execução e são necessárias para garantir a semântica correta entre instruções dependentes de dados. Além disso, o DCE continua produzindo réplicas até que o desvio que gerou os fluxos seja resolvido. Assim, uma seção completa do código pode ser replicado, reduzindo o desempenho. Uma alternativa natural para esse problema é reusar essas seções (ou traços) que são replicadas. O objetivo desse trabalho é analisar e avaliar a efetividade do reuso de valores na arquitetura DCE. Como será apresentado, o princípio do reuso, em diferentes granularidades, pode reduzir efetivamente o problema das réplicas e levar a aumentos de desempenho. / The Dynamic Conditional Execution (DCE) is an alternative to reduce the cost of mispredicted branches. The basic idea is to fetch all paths produced by a branch that obey certain restrictions regarding complexity and size. As a consequence, a smaller number of predictions is performed, and therefore, a lower number branches is mispredicted. Nevertheless, as other multipath solutions, DCE requires a more complex control engine. In a DCE architecture, one may observe that several replicas of the same instruction are dispatched to the functional units, blocking resources that might be used by other instructions. Those replicas are produced after the join point of the paths and are required to guarantee the correct semantic among data dependent instructions. Moreover, DCE continues producing replicas until the branch that generated the paths is resolved. Thus, a whole section of code may be replicated, harming performance. A natural alternative to this problem is the attempt to reuse those replicated sections, namely the replicated traces. The goal of this work is to analyze and evaluate the effectiveness of value reuse in DCE architecture. As it will be presented, the principIe of reuse, in different granularities, can reduce effectively the replica problem and lead to performance improvements.

A indústria de mineração investiga continuamente processos de reconciliação e técnicas adequadas para mapear possíveis riscos na recuperação do minério e no planejamento de lavra. Um estudo de caso em uma mina de cobre brasileira investiga a adequação do uso de teores simulados para a definição de áreas de risco que afetam o planejamento mineiro e as reservas minerais definidas. Simulações condicionais foram usadas para derivar múltiplos modelos de teores de cobre dentro de um típico corpo de minério do depósito e esses modelos foram comparados com os dados reais de produção (reconciliação). A comparação permitiu uma melhor compreensão sobre a variabilidade da qualidade e ajudou na definição de um plano de produção mais consistente. Com os intervalos de valores provenientes das simulações foi possível mapear importantes áreas de incerteza que afetam o plano de lavra. Análises de risco foram conduzidas para a definição da cava final através da comparação do limite previsto pela estimativa por krigagem ordinária e as cavas resultantes dos modelos simulados de teor de cobre, sendo possível avaliar os impactos operacionais nos principais fatores econômicos (valor presente líquido e fluxo de caixa descontado). Várias cavas finais foram geradas para vários modelos de blocos simulados, comparando-se com a cava anteriormente planejada através da utilização do modelo de blocos produzido pela krigagem ordinária. Além disso, foi definida uma escala de risco utilizando os modelos condicionalmente simulados e os valores estimados pela krigagem (comumente utilizados como referência para definição do orçamento e da alimentação de metal no moinho) a fim de minimizar as variações dos teores. Estas medidas se propõem a reduzir o risco no cumprimento das metas de produção de metal no planejamento de lavra. Esta abordagem, considerando o risco durante o sequenciamento de lavra, foi utilizada no plano de produção anual da mina e comparada com o realizado no período e com a abordagem tradicionalmente aplicada. Os resultados evidenciaram possíveis riscos associados com o limite da cava final operacional e demonstraram os benefícios do uso de análise de risco como uma ferramenta para visualizar e analisar os limites de cava final e planejamento de lavra, ajudando a tomar melhores decisões estratégicas no gerenciamento da mina referente aos fatores técnicos e indicadores econômicos. / Mining industry continuously investigates proper reconciliation procedures and techniques for mapping possible risks in ore recovery and mine planning. A case study at a Brazilian copper mine investigates the adequacy of using simulated grades for defining high risk areas affecting mine planning and the defined mineral reserves. Conditional simulations were used to derive multiple copper grade models within a typical ore-body of the deposit and compared these models against the real production data (reconciliation). The comparison allowed a better understanding of grade variability and helped in defining a more consistent mine plan. The range of values derived from the simulations mapped areas of significant uncertainty affecting the pushback’s plan. Risk analysis were conducted for the final pit definition by comparing the projected final pit limit and mining sequence against the simulated models of copper grade, assessing the operational impacts on key economical factors (mine net present value, discounted cash flow). Several final pit limits were generated for various simulated grade block models and compared to the previously pit planned using the ordinary kriging grade block model. Also, it was defined a risk scale using the conditional simulated models and the ordinary kriged values (commonly used for budget and mill feed reference) in order to minimize grade variations ensuring less risk on completion of the metal production goals on mine planning. This approach considering the risk for the mining sequencing was used on the annual mining plan and compared to the real production and to the traditional approach. The results highlighted possible risks associated with the operational pit limit and demonstrated the benefits of using risk analysis as a tool to preview and review the final pit limits and mine planning, helping to make better strategic decisions by the copper mine management related to the technical factors and economical indicators.

This master thesis concerns itself with the alternative procedural measure, the institute of conditional discontinuance of criminal prosecution and some of its relating issues. Conditional discontinuance of criminal prosecution is a type of diversion. It is employed by public prosecutors, or courts, to suspend the further criminal prosecution of those who are accused of committing less serious criminal offence. This is carried on the condition that the accused individual is presented with the opportunity to observe the behavioural rules of proper life throughout the probationary period. Should there be proof of the individual's observation of the rules, the decision regarding his/her case becomes final. However, should the accused fail to observe such rules during the probationary period, then the public prosecutor, or the court, for that matter, rule in favour of furthering his/her prosecution (albeit this rarely takes place in practice). Conditional discontinuance of criminal prosecution is a type of diversion designed to reduce the burden on law enforcement authorities. It is based on the idea of restorative justice with an emphasis on the victim and his/her interests. The primary objective is to bring the accused to the victim and compensate the damages. Before the ruling, it is not necessary...

Skinner (1957) attested that the acquisition of one type of verbal operant will not necessarily occasion the emergence of another type of verbal response topography. In contrast, several studies have shown that multiple exemplar training (MET) is a mechanism that can facilitate the emergence of untrained operants, and it has been considered a powerful tool for establishing generalized operant responses also known as derived relational responses in the language of Relational Frame Theory (RFT). Using a multiple probe design across participants, the current study evaluated the effects of two training protocols in the emergence of untaught intraverbal responses (listing and vocal spelling of words). In Experiment 1, four participants diagnosed with intellectual disability were trained in taking dictation responses and tested for the emergence of intraverbal responses in the form of vocal spelling of words. In Experiment 2, three out of the four participants were trained to relate three sets of three synonyms each using a conditional discrimination training. The results demonstrated that the training procedures used during both experiments were effective in occasioning the emergence of untrained intraverbal responses. It was suggested that participants should have had a history of MET through the course of their academic life which facilitated the emergence of different intraverbal responses in this study.

Neural networks based deep learning methods have gained significant success in several real world tasks: from machine translation to web recommendation, and it is also greatly improving the computer vision and the natural language processing. Compared with conventional machine learning techniques, neural network based deep learning do not require careful engineering and consideration domain expertise to design a feature extractor that transformed the raw data to a suitable internal representation. Its extreme efficacy on multiple levels of representation and feature learning ensures this type of approaches can process high dimensional data. It integrates the feature representation, learning and recognition into a systematical framework, which allows the learning starts at one level (i.e., being with raw input) and end at a higher slightly more abstract level. By simply stacking enough such transformations, very complex functions can be obtained. In general, high level feature representation facilitate the discrimination of patterns, and additionally can reduce the impact of irrelevant variations. However, previous studies indicate that deep composition of the networks make the training errors become vanished. To overcome this weakness, several techniques have been developed, for instance, dropout, stochastic gradient decent and residual network structures. In this study, we incorporates latent information into different network structures (e.g., restricted Boltzmann machine, recursive neural networks, and long short term memory). The conditional latent information reflects the high dimensional correlation existed in the data structure, and the typical network structure may not learn this kind of features due to limitation of the initial design (i.e., the network size the parameters). Similarly to residual nets, the conditional neural networks jointly learns the global features and local features, and the specifically designed network structure helps to incorporate the modulation derived from the probability distribution. The proposed models have been widely tested in different datasets, for instance, the conditional RBM has been applied to detect the speech components, and a language model based gated RBM has been used to recognize speech related EEG patterns. The conditional RNN has been tested in both general natural language modeling and medical notes prediction tasks. The results indicate that by introducing conditional branches in the conventional network structures, the latent features can be globally and locally learned.

x, 94 p. : ill. (some col.) / This paper examines how competition faced by firms affects asset risk and expected returns. Contrary to Hou and Robinson's (2006) findings, I find that cross-industry variation in competition, as measured by the concentration ratio, is not a robust determinant of unconditional expected stock returns. In contrast, within-industry competition, as measured by relative price markup, is positively related to expected stock returns. Moreover, this relation is not captured by commonly used models of expected returns. When using the Markov regime-switching model advocated by Perez-Quiros and Timmermann (2000), I test and find support for Aguerrevere's (2009) recent model of competition find risk dynamics. In particular, systematic risk is greater in more competitive industries during bad times and greater in more concentrated industries during good times. In addition, real investment by firms facing greater competition leads real investment by firms facing less competition, supporting Aguerrevere's notion that less competition results in higher growth options and hence higher risk in good times. / Committee in charge: Dr. Roberto Gutierrez, Chair; Dr. Roberto Gutierrez, Advisor; Dr. Diane Del Guercio, Inside Member; Dr. John Chalmers， Inside Member； Dr. Bruce Blonigen， Outside Member

Dans le cadre de cette thèse, nous présentons l’extension des résultats sur le raisonnement avec des bases de croyances totalement ordonnées au cas partiellement ordonné. L’idée est de raisonner avec des bases logiques équipées d’un ordre partiel exprimant la certitude relative et de construire une fermeture déductive partiellement ordonnée. Au niveau syntaxique, nous pouvons soit utiliser un langage exprimant des paires de formules et des axiomes décrivant les propriétés de l’ordre, ou utiliser des formules en relation avec des poids symboliques partiellement ordonnés dans l’esprit de la logique possibiliste. Une sémantique possible consiste à supposer que cet ordre provient d’un ordre partiel sur les modèles. Elle exige la capacité d’induire un ordre partiel sur les sous-ensembles d’un ensemble, à partir d’un ordre partiel sur ses éléments. Parmi plusieurs définitions de relations d’ordre partiel ainsi définies, nous sélectionnons la plus pertinente pour représenter la notion de certitude relative, en accord avec la théorie des possibilités. Nous montrons les limites d’une sémantique basée sur un ordre partiel unique sur les modèles et proposons une sémantique plus générale qui utilise une relation d’ordre partiel entre les ensembles de modèles. Nous utilisons un langage de plus haut niveau qui exprime des conjonctions de paires de formules en relation, avec des axiomes qui décrivent les propriétés de la relation. Nous proposons deux approches syntaxiques pour inférer de nouvelles paires de formules à partir d’une base partiellement ordonnée, et compléter ainsi l’ordre sur le langage propositionnel. L’une des inférences est proche des logiques conditionnelles de Lewis (qui traite le cas totalement ordonné) et d’un travail de Halpern. Elle est également proche du Système P. Nous reprenons la logique possibiliste symbolique proposée par Benferhat et Prade et comparons cette approche avec l’approche par certitude relative. Pour cela nous poursuivons l’étude de la logique possibiliste symbolique en démontrant un résultat de complétude. Nous étudions la question de la traduction d’une base partiellement ordonnée en base possibiliste symbolique et inversement. Nous proposons enfin des pistes pour une implémentation du système d’inférence de certitude relative et du système possibiliste symbolique. / In this thesis, we present results on the extension of the existing methods for reasoning with totally ordered belief bases to the partially ordered case. The idea is to reason from logical bases equipped with a partial order expressing relative certainty and to construct a partially ordered deductive closure. The difficult part lies in the fact that equivalent definitions in the totally ordered case are no longer equivalent in the partially ordered case. At the syntactic level we can either use a language expressing pairs of related formulas and axioms describing the properties of the ordering, or use formulas with partially ordered symbolic weights attached to them in the spirit of possibilistic logic. A possible semantics consists in assuming that the partial order on formulas stems from a partial order on interpretations. It requires the capability of inducing a partial order on subsets of a set from a partial order on its elements so as to extend possibility theory functions. Among different possible definitions of induced partial order relations, we select the one generalizing necessity orderings (closely related to epistemic entrenchments). We study such a semantic approach inspired from possibilistic logic, and show its limitations when relying on a unique partial order on interpretations. We propose a more general sound and complete approach to relative certainty, inspired by conditional modal logics, in order to get a partial order on the whole propositional language. Some links between our approach and several inference systems, namely conditional logic, modal epistemic logic and non-monotonic preferential inference are established. Possibilistic logic with partially ordered symbolic weights proposed by Benferhat and Prade is also revisited and we continue the study by proving a completeness result. A comparison with the relative certainty approach is made via mutual translations. We compare this approach with the relative certainty approach.We study the question of the translation of a partially ordered base into a symbolic possibilistic base and vice versa. The results for this translation highlight different assumptions underlying the two logics. We also offer steps toward implementation tools for the inference of relative certainty and for the symbolic possibilistic system.

Objetivo: Apresentar e discutir métodos para estimar prevalências ajustadas em pesquisas clínicas e epidemiológicas, bem como desenvolver rotinas computacionais em SAS e R. Métodos: No contexto de estudo transversal, foi simulada uma amostra de 2.000 observações independentes, considerando o desfecho dicotômico diabetes, sexo como a variável de exposição e idade como variável de ajuste. As estimativas de prevalências ajustadas (IC 95%) foram estimadas pelos métodos de predição condicional e marginal, utilizando as rotinas desenvolvidas em SAS e R. O método Delta foi usado para construir os intervalos de confiança. Os resultados foram comparados com aqueles do SUDAAN (SAS-Callable), Stata e a macro %ADJ_PROP (SAS). Resultados: No exemplo simulado, 68,2% são do sexo feminino e a idade média (DP) foi 57,6 (5,0) anos, sendo 54,2 (3,9) anos em homens e 59,2 (4,6) anos em mulheres. A estimativa da prevalência global do desfecho foi de 25,3% (IC 95%:23,4-27,3); sendo 13,8% (IC 95%:11,7-16,7) e 30,7% (IC 95%:28,3-33,2), respectivamente para homens e mulheres. As estimativas de prevalências ajustadas por idade, por meio do método de predição condicional, foram de 19,6% (IC 95%:16,2-23,6) para homens, e 23,6% (IC 95%:21,2-26,1) para mulheres. Pelo método de predição marginal, as estimativas foram de 22,4% (IC 95%:18,7-26,5) para homens, e 26,3% (IC 95%:24,1-28,6) para mulheres. Conclusão: A discrepância entre as estimativas não ajustadas é devida ao confundimento pela idade. Estimativas livres de confundimento podem ser obtidas por meio das prevalências ajustadas pela idade. No entanto, a estimativa pelo método de predição condicional não engloba a prevalência global. Em virtude disso, o método de predição marginal é, geralmente, mais adequado. A rotina desenvolvida na versão para R é uma alternativa aos softwares comerciais. / Objective: To present and discuss methods to estimate adjusted prevalences for clinical and epidemiological research, and develop computational routines in SAS and R. Methods: In the context of cross-sectional study, it was simulated a sample of 2,000 independent observations, considering the dichotomous outcome diabetes, sex as the exposure variable and age as an adjustment variable. Adjusted prevalences were estimated by the conditional and marginal methods, using routines developed in SAS and R. Confidence intervals were constructed using the Delta method. The results were compared with those of the SUDAAN (SAS-callable), Stata and macro %ADJ_PROP (SAS). Results: In simulated example, 68.2% are female and the mean (SD) age was 57.6 (5.00) years old, being that 54.2 (3.94) years for men and 59.2 (4.60) years in women. The estimated global prevalence of outcome was 25.3% (CI 95%: 23.4-27.3) and 13.8% (CI 95%: 11.7-16.7) and 30.7% (CI 95%: 28.3-33.2), respectively for men and women. Estimates of adjusted prevalence for age, through the conditional method, were 19.6% (CI 95%: 16.2-23.6) for men, and 23.6% (CI 95%: 21,2-26.1) for women. For marginal method, the estimates were 22.4% (CI 95%: 18.7-26.5) for men and 26.3% (CI 95%: 24.1-28.6) for women. Conclusion: The observed discrepancy in estimates by sex, unadjusted, can be attributed to confounding due to difference in age distribution between sexes. Comparable estimates (without confounding) of the prevalences can be obtained through prevalence adjusted for age. However, the estimate for the conditional method does not comprise the global prevalence. As a result, the marginal method is in general more suitable. The developed routines can be useful for estimating adjusted prevalences, particularly the R version (an alternative to commercial software).