[pt] O objetivo da dissertação é analisar e comparar a influência de diferentes tipos de parceiros em arranjos cooperativos para projetos de PD&I sobre a apropriação de valor pelas empresas inovadoras da indústria de transformação no Brasil, considerando-se condições ambientais distintas, como nível de intensidade tecnológica e força do regime de apropriabilidade dos setores em que atuam, bem como características internas, como tamanho da empresa e capacidade para cooperar em projetos de PDeI (focalizando inovações de produto e/ou processo). A fonte de dados é a Pesquisa Nacional de Inovação (Pintec 2014), realizada pelo Instituto Brasileiro de Geografia e Estatística (IBGE). A pesquisa pode ser considerada descritiva e aplicada. A metodologia adotada compreende pesquisa bibliográfica sobre inovação; mecanismos de apropriação de valor; cooperação interorganizacional para inovação e classificações tecnológicas, destacando-se a classificação de intensidade tecnológica proposta pela Organização para a Cooperação e Desenvolvimento Econômico (OCDE); pesquisa documental referente à Classificação CNAE e ao Manual da Pintec 2014, ambas divulgadas pelo IBGE; análise de conteúdo para classificar as atividades econômicas das empresas inovadoras da indústria de transformação (respondentes da Pintec 2014), segundo quatro níveis de intensidade tecnológica e regime de apropriabilidade dos setores em que atuam e três faixas de pessoal alocado; solicitação ao IBGE de acesso aos microdados não desidentificados da Pintec 2014; e desenvolvimento de modelos econométricos logit para os doze agrupamentos de empresas, classificadas por intensidade tecnológica/força do regime de apropriabilidade do setor e por faixa de pessoal ocupado. A utilização dos microdados da Pintec 2014 para analisar a influência da cooperação interorganizacional sobre a apropriação de valor pelas empresas inovadoras da indústria de transformação no Brasil em diferentes condições ambientais conferem à pesquisa um caráter original, uma vez que os estudos anteriores baseados em Pesquisas Nacionais de Inovação não exploraram essa abordagem metodológica. / [en] This dissertation aims to analyze and compare the influence of different types of partners in cooperative arrangements for RDandI projects on the appropriation of value by innovative companies of the transformation industry in Brazil, considering different environmental conditions, such as the level of technological intensity and strength of the appropriability regime of the sectors in which they operate, as well as internal characteristics such as company size and its capacity to cooperate in RDandI projects (focusing on product and/or process innovations). The primary data source is the National Innovation Survey (Pintec), conducted by the Brazilian Institute of Geography and Statistics (IBGE). The research can be considered descriptive and applied. The methodology adopted includes bibliographic research on innovation; value creation and value appropriation mechanisms; and technological classifications, highlighting the classification of sectoral technological intensity proposed by the Organization for Economic Cooperation and Development (OECD); documentary analysis concerning the Brazilian Classification of Economic Activities (CNAE) and Pintec 2014, both published by IBGE; content analysis to classify the economic activities of the respondent companies, according to sectoral technological intensity, appropriability regime of the sectors in which they operate and number of employees; request to IBGE for access to microdata of Pintec 2014; development of logit econometric models for the companies classified by sector technological intensity/strength of the appropriability regime, and by number of employees (micro and small, medium and large companies). The use of microdata from Pintec 2014 to analyze and compare the influence of different types of partners in cooperative arrangements for RDandI projects on the appropriation of value by innovative companies under different environmental conditions give the research an original character, since previous studies based on National Innovation Surveys have not explored this methodological approach.

Cette thèse porte sur les déterminants de la migration des compétences au Liban, dont l'objet principal est l'identification des facteurs et des risques associés à la migration chez les individus hautement qualifiés. La problématique se résume par la question suivante : À formation universitaire égale, quel diplômé devient-il migrant ? Afin de combler le manque de données sur le sujet, nous avons réalisé une enquête (en ligne) auprès des diplômés de la Lebanese American University et de l'Université Saint-Esprit de Kaslik, au Liban. Ces diplômés forment une pluralité et une mixité culturelle, linguistique et socio-économique représentatives des étudiants du Liban. L'objectif de l'enquête était de comparer les similarités et de contraster les différences entre des diplômés migrants et non-migrants, afin de comprendre les logiques différenciées de leurs comportements migratoires. Au début, nous avons effectué une segmentation des diplômés pour les répartir en sous-groupes en fonction de leur statut migratoire, ce qui nous a permis d'identifier et de comprendre les logiques différenciées de leurs comportements migratoires. L'analyse descriptive des résultats de l'enquête révèle des différences en termes de caractéristiques démographiques, économiques et familiales entre diplômés migrants et non-migrants, mais un peu moins de divergences en ce qui concerne leurs parcours universitaires et le domaine des études. À partir de la modélisation, nous avons démontré comment le risque de migrer à l'étranger peut être déterminé par certains facteurs individuels et familiaux, mais surtout en fonction du temps, c'est-à-dire en fonction de la durée depuis l'obtention du diplôme universitaire le plus élevé. Nous démontrons ainsi que, même à formation universitaire égale, le capital humain et le capital social peuvent engendrer des migrations internationales parmi des diplômés ayant effectué un même parcours universitaire et ayant vécu les mêmes conditions socio-économiques au Liban. Nous examinons également des obstacles qui freinent la migration des compétences, pour finalement analyser brièvement les facteurs qui déterminent la migration de retour au Liban. Sur la base de ces conclusions, nous confirmons que nous avons vérifié nos hypothèses par les faits qui se basent sur les résultats de notre enquête. Malgré la difficulté relative à l'étude des migrations internationales en générale et à l'utilisation de la technique de l'enquête en ligne, nous avons réussi à obtenir des résultats très intéressants, que nous avons comparés à des données disponibles sur la migration des compétences au Liban et à l'échelle globale. / This thesis focuses on the determinants of the highly skilled migration in Lebanon whose main purpose is to identify factors and risks associated with migration among highly skilled graduates. The research problem is summarized in the following main question: Given equal level of education, which graduate becomes a migrant? To address the lack of data on this particular topic, we conducted a (online) survey on graduates from the Lebanese American University and the Holy Spirit University of Kaslik in Lebanon. Graduates from both universities form diverse cultural, linguistic and socioeconomic characteristics that are representative of Lebanese graduates in general. The aim of the survey was to compare similarities and contrast differences between migrant and non-migrant graduates in order to understand their diverse behavior with regard to migration. Based on survey results, we carried out a segmentation of graduates and divided them into sub-groups based on their migration status allowing us to understand their behavior with regard to migration. The descriptive analysis of the survey results reveals differences in demographic, economic and family characteristics between migrant and non-migrant graduates, but little divergence were found regarding their university studies and the field of education. We also modeled a number of risks associated with migration and we demonstrated that the risk of migrating could be determined by a number of individual and family factors, but mainly over time, i.e. the time since graduation with the highest university degree. We also demonstrate that even at equal level of education, human capital and social capital can determine international migration among graduates who have obtained the same level of education and experienced the same socio-economic conditions in Lebanon. We have also explored barriers that hinder migration among skilled graduates and briefly analyzed the main factors determining their return migration to Lebanon. Based on these findings we confirm that we have validated our assumptions by facts based on survey results. Despite the relative difficulty in the study of international migration in general and the implementation of online surveys, we managed to obtain very interesting results which we also compared to available data on skilled migration in Lebanon and at the global level.

Submitted by Rafael Santana (rafael.silvasantana@ufpe.br) on 2017-04-26T17:07:37Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Carolina Marchant.pdf: 5792192 bytes, checksum: adbd82c79b286d2fe2470b7955e6a9ed (MD5) / Made available in DSpace on 2017-04-26T17:07:38Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Carolina Marchant.pdf: 5792192 bytes, checksum: adbd82c79b286d2fe2470b7955e6a9ed (MD5) Previous issue date: 2016-10-31 / CAPES; BOLSA DO CHILE. / In the last decades, univariate Birnbaum-Saunders models have received considerable attention in the literature. These models have been widely studied and applied to fatigue, but they have also been applied to other areas of the knowledge. In such areas, it is often necessary to model several variables simultaneously. If these variables are correlated, individual analyses for each variable can lead to erroneous results. Multivariate regression models are a useful tool of the multivariate analysis, which takes into account the correlation between variables. In addition, diagnostic analysis is an important aspect to be considered in the statistical modeling. Furthermore, multivariate quality control charts are powerful and simple visual tools to determine whether a multivariate process is in control or out of control. A multivariate control chart shows how several variables jointly affect a process. First, we propose, derive and characterize multivariate generalized logarithmic Birnbaum-Saunders distributions. Also, we propose new multivariate generalized Birnbaum-Saunders regression models. We use the method of maximum likelihood estimation to estimate their parameters through the expectation-maximization algorithm. We carry out a simulation study to evaluate the performance of the corresponding estimators based on the Monte Carlo method. We validate the proposed models with a regression analysis of real-world multivariate fatigue data. Second, we conduct a diagnostic analysis for multivariate generalized Birnbaum-Saunders regression models. We consider the Mahalanobis distance as a global influence measure to detect multivariate outliers and use it for evaluating the adequacy of the distributional assumption. Moreover, we consider the local influence method and study how a perturbation may impact on the estimation of model parameters. We implement the obtained results in the R software, which are illustrated with real-world multivariate biomaterials data. Third and finally, we develop a robust methodology based on multivariate quality control charts for generalized Birnbaum-Saunders distributions with the Hotelling statistic. We use the parametric bootstrap method to obtain the distribution of this statistic. A Monte Carlo simulation study is conducted to evaluate the proposed methodology, which reports its performance to provide earlier alerts of out-of-control conditions. An illustration with air quality real-world data of Santiago-Chile is provided. This illustration shows that the proposed methodology can be useful for alerting episodes of extreme air pollution. / Nas últimas décadas, o modelo Birnbaum-Saunders univariado recebeu considerável atenção na literatura. Esse modelo tem sido amplamente estudado e aplicado inicialmente à modelagem de fadiga de materiais. Com o passar dos anos surgiram trabalhos com aplicações em outras áreas do conhecimento. Em muitas das aplicações é necessário modelar diversas variáveis simultaneamente incorporando a correlação entre elas. Os modelos de regressão multivariados são uma ferramenta útil de análise multivariada, que leva em conta a correlação entre as variáveis de resposta. A análise de diagnóstico é um aspecto importante a ser considerado no modelo estatístico e verifica as suposições adotadas como também sua sensibilidade. Além disso, os gráficos de controle de qualidade multivariados são ferramentas visuais eficientes e simples para determinar se um processo multivariado está ou não fora de controle. Este gráfico mostra como diversas variáveis afetam conjuntamente um processo. Primeiro, propomos, derivamos e caracterizamos as distribuições Birnbaum-Saunders generalizadas logarítmicas multivariadas. Em seguida, propomos um modelo de regressão Birnbaum-Saunders generalizado multivariado. Métodos para estimação dos parâmetros do modelo, tal como o método de máxima verossimilhança baseado no algoritmo EM, foram desenvolvidos. Estudos de simulação de Monte Carlo foram realizados para avaliar o desempenho dos estimadores propostos. Segundo, realizamos uma análise de diagnóstico para modelos de regressão Birnbaum-Saunders generalizados multivariados. Consideramos a distância de Mahalanobis como medida de influência global de detecção de outliers multivariados utilizando-a para avaliar a adequacidade do modelo. Além disso, desenvolvemos medidas de diagnósticos baseadas em influência local sob alguns esquemas de perturbações. Implementamos a metodologia apresentada no software R, e ilustramos com dados reais multivariados de biomateriais. Terceiro, e finalmente, desenvolvemos uma metodologia robusta baseada em gráficos de controle de qualidade multivariados para a distribuição Birnbaum-Saunders generalizada usando a estatística de Hotelling. Baseado no método bootstrap paramétrico encontramos aproximações da distribuição desta estatística e obtivemos limites de controle para o gráfico proposto. Realizamos um estudo de simulação de Monte Carlo para avaliar a metodologia proposta indicando seu bom desempenho para fornecer alertas precoces de processos fora de controle. Uma ilustração com dados reais de qualidade do ar de Santiago-Chile é fornecida. Essa ilustração mostra que a metodologia proposta pode ser útil para alertar sobre episódios de poluição extrema do ar, evitando efeitos adversos na saúde humana.

Dans cette thèse, nous étudions des modèles définis par des équations de moments conditionnels. Une grande partie de modèles statistiques (régressions, régressions quantiles, modèles de transformations, modèles à variables instrumentales, etc.) peuvent se définir sous cette forme. Nous nous intéressons au cas des modèles avec un paramètre à estimer de dimension finie, ainsi qu’au cas des modèles semi paramétriques nécessitant l’estimation d’un paramètre de dimension finie et d’un paramètre de dimension infinie. Dans la classe des modèles semi paramétriques étudiés, nous nous concentrons sur les modèles à direction révélatrice unique qui réalisent un compromis entre une modélisation paramétrique simple et précise, mais trop rigide et donc exposée à une erreur de modèle, et l’estimation non paramétrique, très flexible mais souffrant du fléau de la dimension. En particulier, nous étudions ces modèles semi paramétriques en présence de censure aléatoire. Le fil conducteur de notre étude est un contraste sous la forme d’une U-statistique, qui permet d’estimer les paramètres inconnus dans des modèles généraux. / In this dissertation we study statistical models defined by condition estimating equations. Many statistical models could be stated under this form (mean regression, quantile regression, transformation models, instrumental variable models, etc.). We consider models with finite dimensional unknown parameter, as well as semiparametric models involving an additional infinite dimensional parameter. In the latter case, we focus on single-index models that realize an appealing compromise between parametric specifications, simple and leading to accurate estimates, but too restrictive and likely misspecified, and the nonparametric approaches, flexible but suffering from the curse of dimensionality. In particular, we study the single-index models in the presence of random censoring. The guiding line of our study is a U-statistics which allows to estimate the unknown parameters in a wide spectrum of models.

[ES] La Ley de Prevención de Riesgos Laborales de 8 de noviembre de 1995 (LPRL), en vigor desde el 10 de febrero de 1996, establece en su artículo 5: "tendrá por objeto la promoción de la mejora de las condiciones de trabajo dirigida a elevar el nivel de protección de la seguridad y la salud de los trabajadores en el trabajo." En esta Tesis se ha investigado la evolución de los índices de siniestralidad laboral y su relación con la evolución de diferentes variables explicativas relacionadas con el desarrollo normativo, el mercado de trabajo, la estructura productiva, las condiciones de empleo y las condiciones individuales, entre otras, para el caso de España y en el periodo 1995-2017, que abarca desde la promulgación de la LPRL hasta fechas recientes donde se disponía de los datos históricos necesarios. La investigación se ha centrado en los índices de salud más relevantes según su significado en términos de riesgo y/o sus componentes. El objetivo de la investigación ha sido el encontrar evidencias sobre relaciones causa-efecto entre índices y variables, a partir de las cuales extraer lecciones que facilitarán una mejor planificación de la acción preventiva. Para ello, se han propuesto varios modelos explicativos utilizando diferentes herramientas estadísticas, que han permitido formular de manera explícita y analizar la relación entre la evolución de los indicadores de salud ocupacional y la evolución de las principales variables explicativas. En términos generales puede concluirse que la implantación de dicha ley y normativa que la acompaña ha tenido un impacto positivo en las condiciones de trabajo y en consecuencia sobre el nivel de seguridad y salud de los trabajadores desde entonces y hasta la fecha. Sin embargo, se observan diferentes comportamientos cíclicos en la evolución de los indicadores, tales como los índices de incidencia, frecuencia y gravedad, que pone de manifiesto su dependencia de la naturaleza y comportamiento cíclico de algunas de las variables explicativas más importantes relacionadas con ciclos económicos, mercado de trabajo, estructura productiva, etc. Además, se observa como aspectos tales como la pertenencia a grupos de edad jóvenes o expertos, el nivel de estudios, determinadas categorías profesionales, y algunos sectores particulares tienen efectos significativos sobre los valores alcanzados por los índices de siniestralidad. En cambio, otros, como el trabajo a tiempo parcial o la contratación temporal no manifiestan tener tanta repercusión sobre los indicadores. / [CA] Partint de les dades corresponents als accidents ocorreguts en el període 1995-2017, es La Llei de Prevenció de Riscos Laborals de 8 de novembre de 1995 (*LPRL), en vigor des del 10 de febrer de 1996, estableix en el seu article 5: "tindrà per objecte la promoció de la millora de les condicions de treball dirigida a elevar el nivell de protecció de la seguretat i la salut dels treballadors en el treball." En aquesta Tesi s'ha investigat l'evolució dels índexs de sinistralitat laboral i la seua relació amb l'evolució de diferents variables explicatives relacionades amb el desenvolupament normatiu, el mercat de treball, l'estructura productiva, les condicions d'ocupació i les condicions individuals, entre altres, per al cas d'Espanya i en el període 1995-2017, que abasta des de la promulgació de la LPRL fins a dates recents on es disposava de les dades històriques necessàries. La investigació s'ha centrat en els índexs de salut més rellevants segons el seu significat en termes de risc i/o els seus components. L'objectiu de la investigació ha sigut el trobar evidències sobre relacions causa-efecte entre índexs i variables, a partir de les quals extraure lliçons que facilitaran una millor planificació de l'acció preventiva. Per a això, s'han proposat diversos models explicatius utilitzant diferents eines estadístiques, que han permés formular de manera explícita i analitzar la relació entre l'evolució dels indicadors de salut ocupacional i l'evolució de les principals variables explicatives. En termes generals pot concloure's que la implantació d'aquesta llei i normativa que l'acompanya ha tingut un impacte positiu en les condicions de treball i en conseqüència sobre el nivell de seguretat i salut dels treballadors des de llavors i fins hui. No obstant això, s'observen diferents comportaments cíclics en l'evolució dels indicadors, com ara els índexs d'incidència, freqüència i gravetat, que posa de manifest la seua dependència de la naturalesa i comportament cíclic d'algunes de les variables explicatives més importants relacionades amb cicles econòmics, mercat de treball, estructura productiva, etc. A més, s'observa com a aspectes com ara la pertinença a grups d'edat joves o experts, el nivell d'estudis, determinades categories professionals, i alguns sectors particulars tenen efectes significatius sobre els valors aconseguits pels índexs de sinistralitat. En canvi, uns altres, com el treball a temps parcial o la contractació temporal no manifesten tindre tanta repercussió sobre els indicadors. / [EN] The Occupational Risk Prevention Act of November 8, 1995 (ORPA), in force since February 10, 1996, establishes in its article 5: "will have as its objective the promotion of the improvement of working conditions aimed at raise the level of protection of the safety and health of workers at work. " This thesis has investigated the evolution of the occupational accident rates and their relationship with the evolution of different explanatory variables related to regulatory development, the labor market, the productive structure, employment conditions and individual conditions, among others, in the case of Spain and in the period 1995-2017, which ranges from the enactment of the LPRL to recent dates where the necessary historical data was available. Research has focused on the most relevant health indices according to their meaning in terms of risk and / or their components. The objective of the research has been to find evidence on cause-effect relationships between indices and variables, from which to extract lessons that will facilitate better planning of preventive action. To this end, several explanatory models have been proposed using different statistical tools, which have made it possible to explicitly formulate and analyze the relationship between the evolution of occupational health indicators and the evolution of the main explanatory variables. In general terms, it can be concluded that the implementation of said law and accompanying regulations has had a positive impact on working conditions and consequently on the level of health and safety of workers since then and to date. However, different cyclical behaviors are observed in the evolution of the indicators, such as incidence, frequency and severity indices, which highlights their dependence on the nature and cyclical behavior of some of the most important explanatory variables related to economic cycles, labor market, productive structure, etc. Furthermore, aspects such as belonging to young age groups or experts, educational level, certain professional categories, and some particular sectors are observed as having significant effects on the values reached by the accident rates. On the other hand, others, such as part-time work or temporary hiring, do not claim to have such an impact on the indicators. / Gallego Blasco, VS. (2021). Análisis de la incidencia de factores causales en la evolución de la siniestralidad laboral en España [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/168774 / TESIS

<p>Sequential decision-making is a fundamental class of problem that motivates algorithm designs of online machine learning and reinforcement learning. Arguably, the resulting online algorithms have supported modern online service industries for their data-driven real-time automated decision making. The applications span across different industries, including dynamic pricing (Marketing), recommendation (Advertising), and dosage finding (Clinical Trial). In this dissertation, we contribute fundamental statistical design advances for sequential decision-making algorithms, leaping progress in theory and application of online learning and sequential decision making under uncertainty including online sparse learning, finite-armed bandits, and high-dimensional online decision making. Our work locates at the intersection of decision-making algorithm designs, online statistical machine learning, and operations research, contributing new algorithms, theory, and insights to diverse fields including optimization, statistics, and machine learning.</p> <p><br></p> <p>In part I, we contribute a theoretical framework of continuous risk monitoring for regularized online statistical learning. Such theoretical framework is desirable for modern online service industries on monitoring deployed model's performance of online machine learning task. In the first project (Chapter 1), we develop continuous risk monitoring for the online Lasso procedure and provide an always-valid algorithm for high-dimensional dynamic pricing problems. In the second project (Chapter 2), we develop continuous risk monitoring for online matrix regression and provide new algorithms for rank-constrained online matrix completion problems. Such theoretical advances are due to our elegant interplay between non-asymptotic martingale concentration theory and regularized online statistical machine learning.</p> <p><br></p> <p>In part II, we contribute a bootstrap-based methodology for finite-armed bandit problems, termed Residual Bootstrap exploration. Such a method opens a possibility to design model-agnostic bandit algorithms without problem-adaptive optimism-engineering and instance-specific prior-tuning. In the first project (Chapter 3), we develop residual bootstrap exploration for multi-armed bandit algorithms and shows its easy generalizability to bandit problems with complex or ambiguous reward structure. In the second project (Chapter 4), we develop a theoretical framework for residual bootstrap exploration in linear bandit with fixed action set. Such methodology advances are due to our development of non-asymptotic theory for the bootstrap procedure.</p> <p><br></p> <p>In part III, we contribute application-driven insights on the exploration-exploitation dilemma for high-dimensional online decision-making problems. Such insights help practitioners to implement effective high-dimensional statistics methods to solve online decisionmaking problems. In the first project (Chapter 5), we develop a bandit sampling scheme for online batch high-dimensional decision making, a practical scenario in interactive marketing, and sequential clinical trials. In the second project (Chapter 6), we develop a bandit sampling scheme for federated online high-dimensional decision-making to maintain data decentralization and perform collaborated decisions. These new insights are due to our new bandit sampling design to address application-driven exploration-exploitation trade-offs effectively. </p>

Background: Shifting the healthcare payment system from a volume-based to a value-based model has been a significant effort to improve the quality of care and reduce healthcare costs in the US. In 2018, Massachusetts Medicaid launched Accountable Care Organizations (ACOs) as part of the effort. Constructing, assessing, and risk-adjusting quality measures are integral parts of the reform process. Methods: Using data from the MassHealth Data Warehouse (2016-2019), we assessed the loss of community tenure (CTloss) as a potential quality measure for patients with bipolar, schizophrenia, or other psychotic disorders (BSP). We evaluated various statistical models for predicting CTloss using deviance, Akaike information criterion, Vuong test, squared correlation and observed vs. expected (O/E) ratios. We also used logistic regression to investigate risk factors that impacted medication nonadherence, another quality measure for patients with bipolar disorders (BD). Results: Mean CTloss was 12.1 (±31.0 SD) days in the study population; it varied greatly across ACOs. For risk adjustment modeling, we recommended the zero-inflated Poisson or doubly augmented beta model. The O/E ratio ranged from 0.4 to 1.2, suggesting variation in quality, after adjusting for differences in patient characteristics for which ACOs served as reflected in E. Almost half (47.7%) of BD patients were nonadherent to second-generation antipsychotics. Patient demographics, medical and mental comorbidities, receiving institutional services like those from the Department of Mental Health, homelessness, and neighborhood socioeconomic stress impacted medication nonadherence. Conclusions: Valid quality measures are essential to value-based payment. Heterogeneity implies the need for risk adjustment. The search for a model type is driven by the non-standard distribution of CTloss.

[pt] Prever com precisão a resposta a drogas para uma dada amostra baseado em características moleculares pode ajudar a otimizar o desenvolvimento de drogas e explicar mecanismos por trás das respostas aos tratamentos. Nessa dissertação, dois estudos de caso foram gerados, cada um aplicando diferentes dados genômicos para a previsão de resposta a drogas. O estudo de caso 1 avaliou dados de perfis de metilação de DNA como um tipo de característica molecular que se sabe ser responsável por causar tumorigênese e modular a resposta a tratamentos. Usando perfis de metilação de 987 linhagens celulares do genoma completo na base de dados Genomics of Drug Sensitivity in Cancer (GDSC), utilizamos algoritmos de aprendizado de máquina para avaliar o potencial preditivo de respostas citotóxicas para oito drogas contra o câncer. Nós comparamos a performance de cinco algoritmos de classificação e quatro algoritmos de regressão representando metodologias diversas, incluindo abordagens tree-, probability-, kernel-, ensemble- e distance-based. Aplicando sub-amostragem artificial em graus variados, essa pesquisa procura avaliar se o treinamento baseado em resultados relativamente extremos geraria melhoria no desempenho. Ao utilizar algoritmos de classificação e de regressão para prever respostas discretas ou contínuas, respectivamente, nós observamos consistentemente excelente desempenho na predição quando os conjuntos de treinamento e teste consistiam em dados de linhagens celulares. Algoritmos de classificação apresentaram melhor desempenho quando nós treinamos os modelos utilizando linhagens celulares com valores de resposta a drogas relativamente extremos, obtendo valores de area-under-the-receiver-operating-characteristic-curve de até 0,97. Os algoritmos de regressão tiveram melhor desempenho quando treinamos os modelos utilizado o intervalo completo de valores de resposta às drogas, apesar da dependência das métricas de desempenho utilizadas. O estudo de caso 2 avaliou dados de RNA-seq, dados estes comumente utilizados no estudo da eficácia de drogas. Aplicando uma abordagem de aprendizado semi-supervisionado, essa pesquisa busca avaliar o impacto da combinação de dados rotulados e não-rotulados para melhorar a predição do modelo. Usando dados rotulados de RNA-seq do genoma completo de uma média de 125 amostras de tumor AML rotuladas da base de dados Beat AML (separados por tipos de droga) e 151 amostras de tumor AML não-rotuladas na base de dados The Cancer Genome Atlas (TCGA), utilizamos uma estrutura de modelo semi-supervisionado para prever respostas citotóxicas para quatro drogas contra câncer. Modelos semi-supervisionados foram gerados, avaliando várias combinações de parâmetros e foram comparados com os algoritmos supervisionados de classificação. / [en] Accurately predicting drug responses for a given sample based on molecular features may help to optimize drug-development pipelines and explain mechanisms behind treatment responses. In this dissertation, two case studies were generated, each applying different genomic data to predict drug response. Case study 1 evaluated DNA methylation profile data as one type of molecular feature that is known to drive tumorigenesis and modulate treatment responses. Using genome-wide, DNA methylation profiles from 987 cell lines in the Genomics of Drug Sensitivity in Cancer (GDSC) database, we used machine-learning algorithms to evaluate the potential to predict cytotoxic responses for eight anti-cancer drugs. We compared the performance of five classification algorithms and four regression algorithms representing diverse methodologies, including tree-, probability-, kernel-, ensemble- and distance-based approaches. By applying artificial subsampling in varying degrees, this research aims to understand whether training based on relatively extreme outcomes would yield improved performance. When using classification or regression algorithms to predict discrete or continuous responses, respectively, we consistently observed excellent predictive performance when the training and test sets consisted of cell-line data. Classification algorithms performed best when we trained the models using cell lines with relatively extreme drug-response values, attaining area-under-the-receiver-operating-characteristic-curve values as high as 0.97. The regression algorithms performed best when we trained the models using the full range of drug-response values, although this depended on the performance metrics we used. Case study 2 evaluated RNA-seq data as one of the most popular molecular data used to study drug efficacy. By applying a semi-supervised learning approach, this research aimed to understand the impact of combining labeled and unlabeled data to improve model prediction. Using genome-wide RNA-seq labeled data from an average of 125 AML tumor samples in the Beat AML database (varying by drug type) and 151 unlabeled AML tumor samples in The Cancer Genome Atlas (TCGA) database, we used a semi-supervised model structure to predict cytotoxic responses for four anti-cancer drugs. Semi-supervised models were generated, while assessing several parameter combinations and were compared against supervised classification algorithms.

Cette thèse a été conçu pour explorer les modèles dynamiques de régression, d’évaluer les inférences statistiques pour l’analyse des données de survie et de fiabilité. Ces modèles de régression dynamiques que nous avons considérés, y compris le modèle des hasards proportionnels paramétriques et celui de la vie accélérée avec les variables qui peut-être dépendent du temps. Nous avons discuté des problèmes suivants dans cette thèse.Nous avons présenté tout d’abord une statistique de test du chi-deux généraliséeY2nquiest adaptative pour les données de survie et fiabilité en présence de trois cas, complètes,censurées à droite et censurées à droite avec les covariables. Nous avons présenté en détailla forme pratique deY2nstatistique en analyse des données de survie. Ensuite, nous avons considéré deux modèles paramétriques très flexibles, d’évaluer les significations statistiques pour ces modèles proposées en utilisantY2nstatistique. Ces modèles incluent du modèle de vie accélérés (AFT) et celui de hasards proportionnels (PH) basés sur la distribution de Hypertabastic. Ces deux modèles sont proposés pour étudier la distribution de l’analyse de la duré de survie en comparaison avec d’autre modèles paramétriques. Nous avons validé ces modèles paramétriques en utilisantY2n. Les études de simulation ont été conçus.Dans le dernier chapitre, nous avons proposé les applications de ces modèles paramétriques à trois données de bio-médicale. Le premier a été fait les données étendues des temps de rémission des patients de leucémie aiguë qui ont été proposées par Freireich et al. sur la comparaison de deux groupes de traitement avec des informations supplémentaires sur les log du blanc du nombre de globules. Elle a montré que le modèle Hypertabastic AFT est un modèle précis pour ces données. Le second a été fait sur l’étude de tumeur cérébrale avec les patients de gliome malin, ont été proposées par Sauerbrei & Schumacher. Elle a montré que le meilleur modèle est Hypertabastic PH à l’ajout de cinq variables de signification. La troisième demande a été faite sur les données de Semenova & Bitukov, à concernant les patients de myélome multiple. Nous n’avons pas proposé un modèle exactement pour ces données. En raison de cela était les intersections de temps de survie.Par conséquent, nous vous conseillons d’utiliser un autre modèle dynamique que le modèle de la Simple Cross-Effect à installer ces données. / This thesis was designed to explore the dynamic regression models, assessing the sta-tistical inference for the survival and reliability data analysis. These dynamic regressionmodels that we have been considered including the parametric proportional hazards andaccelerated failure time models contain the possibly time-dependent covariates. We dis-cussed the following problems in this thesis.At first, we presented a generalized chi-squared test statisticsY2nthat is a convenient tofit the survival and reliability data analysis in presence of three cases: complete, censoredand censored with covariates. We described in detail the theory and the mechanism to usedofY2ntest statistic in the survival and reliability data analysis. Next, we considered theflexible parametric models, evaluating the statistical significance of them by usingY2nandlog-likelihood test statistics. These parametric models include the accelerated failure time(AFT) and a proportional hazards (PH) models based on the Hypertabastic distribution.These two models are proposed to investigate the distribution of the survival and reliabilitydata in comparison with some other parametric models. The simulation studies were de-signed, to demonstrate the asymptotically normally distributed of the maximum likelihood estimators of Hypertabastic’s parameter, to validate of the asymptotically property of Y2n test statistic for Hypertabastic distribution when the right censoring probability equal 0% and 20%.n the last chapter, we applied those two parametric models above to three scenes ofthe real-life data. The first one was done the data set given by Freireich et al. on thecomparison of two treatment groups with additional information about log white blood cellcount, to test the ability of a therapy to prolong the remission times of the acute leukemiapatients. It showed that Hypertabastic AFT model is an accurate model for this dataset.The second one was done on the brain tumour study with malignant glioma patients, givenby Sauerbrei & Schumacher. It showed that the best model is Hypertabastic PH onadding five significance covariates. The third application was done on the data set given by Semenova & Bitukov on the survival times of the multiple myeloma patients. We did not propose an exactly model for this dataset. Because of that was an existing oneintersection of survival times. We, therefore, suggest fitting other dynamic model as SimpleCross-Effect model for this dataset.