• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 1
  • 1
  • Tagged with
  • 3
  • 3
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Applications of Monte Carlo Methods in Statistical Inference Using Regression Analysis

Huh, Ji Young 01 January 2015 (has links)
This paper studies the use of Monte Carlo simulation techniques in the field of econometrics, specifically statistical inference. First, I examine several estimators by deriving properties explicitly and generate their distributions through simulations. Here, simulations are used to illustrate and support the analytical results. Then, I look at test statistics where derivations are costly because of the sensitivity of their critical values to the data generating processes. Simulations here establish significance and necessity for drawing statistical inference. Overall, the paper examines when and how simulations are needed in studying econometric theories.
2

Estimação de valores estatísticos de referência para as internações por condições sensíveis a atenção primária em municípios brasileiros

Cavalcanti, Gilmara Alves 22 March 2017 (has links)
Submitted by Viviane Lima da Cunha (viviane@biblioteca.ufpb.br) on 2017-06-26T11:50:59Z No. of bitstreams: 1 arquivototal.pdf: 7515806 bytes, checksum: a09d41d012f6b4bd6ac4e7f7ffd37975 (MD5) / Made available in DSpace on 2017-06-26T11:50:59Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 7515806 bytes, checksum: a09d41d012f6b4bd6ac4e7f7ffd37975 (MD5) Previous issue date: 2017-03-22 / Many indicators have been used to verify the e ciency and quality of services in primary health care. Recently, this level of attention has been evaluated in Brazil. The ambulatory cary sensitive conditions (ACSC), represent a set of health problems that should not reach tertiary care in large numbers, since an e ective primary care action would solve some of these pathologies. Hear, the present study is justi ed by the need to use hospitalizations for ACSC as a base for quality and access indicator for the Primary Care of the health system of Brazilian cities, in association with the Family Health Strategy (FHS). Therefore, we aim to estimate statistical reference values for ambulatory cary sensitive conditions, based on the adjustment of continuous probability distributions, in order to classify the health system of the Brazilian cities into levels of performance, namely: desirable, intermediate, worrying and not acceptable. The thesis that is defended is that the achievement of a statistical reference value, capable of classifying by ASCS, in performance levels for the Brazilian cities, will aid in decision-making processes in the units of FHS, optimizing the health actions in these units, as well as, providing the prevention of unnecessary expenses in the other levels of attention. This is an epidemiological, ecological and inferential study of quantitative approaches, based on the secondary database of the Hospital Information System of the Unique Health System, with the selection of absolute frequencies of diagnostic groups referring to ambulatory cary sensitive conditions to Primary Care, for 2015. Subsequently, these frequencies were transformed into rates per 10,000 inhabitants, making it possible to compare the ACSCs of all Brazilian cities, regardless of their population contingents. From the statistical point of view, the adjustment of continuous probability distributions through the goodness-of- t tests and graphic support were processed using software R, version 3.0.3. Based on the results, a greater incidence of hospital admissions was observed due to pathologies related to gastroenteric, cardiac and renal problems. In view of the three scenarios investigated (Brazil, Northeast and Para ba), the Northeast presents records of hospitalizations for asthma, a diagnosis not observed as one of the three highest incidences at national and state level. It was also evidenced that most of the Brazilian cities investigated in this study have their health systems classi ed as not acceptable. The city of S~ao Paulo, for example, is one such case accounting for 9.471 hospitalizations of patients with heart failure in 2015. Thus, its health system would be considered desirable if there were up to 1.532 records. An intermediate level of performance would range from more than 1.532 to a maximum of 2.944 cases, while the level of concern would be more than 2.944 to 4.177 hospitalizations. An unacceptable performance level is recorded from 4.177 ICSAP. In Brazil, when it comes to the Basic Attention scenario, although advances are perceived, the prospect of reducing the ACSC to minimum values, as is desirable, is still far from the reality. The use of the statistical reference value may constitute a potential tool, ready to use of the administrative managers in order to direct and subsidize proposals of health actions in the cities under their responsibility. / Uma variedade de indicadores tem sido utilizada para averiguar a e ciência e a qualidade de serviços na atenção primaria a saúde. Nos últimos anos, esse nível de atenção vem passando por avaliações no Brasil. As internações por condições sensíveis a atenção primária (ICSAP), representam um conjunto de agravos de saúde que não deveriam chegar em grande quantidade a atenção terciária, posto que uma efetiva ação da atenção primária solucionaria parte dessas patologias. Nesse contexto, a presente pesquisa se justiça pela necessidade de utilizar as ICSAP como base para um indicador de acesso e de qualidade da Atenção básica do sistema de saúde dos municípios brasileiros, em associação com a Estratégia Saúde da Família (ESF). Portanto, objetiva-se estimar valores estatísticos de referência para as internações por condições sensíveis a atenção primária, a partir do ajuste de distribuições de probabilidade contínuas, de modo a classificar o sistema de saúde dos municípios brasileiros em níveis de desempenho, a saber: desejável, intermediário, preocupante e não aceitável. Defende-se a tese de que a obtenção de um valor estatístico de referência, capaz de classificar as ICSAP, em níveis de desempenho para os municípios brasileiros, auxiliará nos processos de tomadas de decisão nas unidades de ESF, otimizando as ações de saúde nessas unidades, bem como, proporcionando a prevenção de gastos desnecessários nos outros níveis de atenção. Trata-se de um estudo epidemiológico, ecológico e inferencial, de abordagem quantitativa, realizado a partir da base de dados secundários do Sistema de Informações Hospitalares do SUS com a seleção das freqüências absolutas dos grupos de diagnósticos referentes as internações por condições sensíveis à atenção primária, referentes ao ano de 2015. A posteriori, as referidas freqüências foram transformadas em taxas por 10.000 habitantes possibilitando a comparação das ICSAP de todos os municípios brasileiros, independente dos seus contingentes populacionais. Sob o ponto de vista estatístico, o ajuste de distribuições de probabilidade contínuas por meio dos testes de aderência e suporte gráfico foram processados utilizando o software R 3.0.3. Com base nos resultados, observou-se uma incidência maior de internações hospitalares devido _a patologias relacionadas a problemas gastroentéricos, cardíacos e renais. Diante dos três cenários investigados (Brasil, Nordeste e Paraíba), a região Nordeste apresenta registros de internações por asma, diagnóstico não observado como uma das três maiores incidências a nível nacional e estadual. Evidenciou-se também, que a maioria dos municípios brasileiros investigados nesse estudo têm seus sistemas de saúde classificados como não aceitáveis. A cidade de São Paulo, por exemplo, é um desses casos, contabilizando 9.471 internações de pacientes com insuficiência cardíaca em 2015. Assim, seu sistema de saúde seria considerado de nível desejável se houvessem até 1.532 registros. Um nível intermediário de desempenho seria de mais de 1.532 até no máximo 2.944 casos, enquanto que, o nível preocupante dar-se-_a com mais de 2.944 até 4.177 internações. Um nível de desempenho não aceitável é registrado a partir de 4.177 ICSAP. No Brasil, em se tratando do quadro de Atenção Básica, embora avanços sejam percebidos, a perspectiva de redução das ICSAP a valores mínimos, como é desejável, ainda está distante da realidade. A utilização do valor estatístico de referência poder á constituir-se de uma potencial ferramenta, a disposição dos gestores administrativos no intuito de direcionar e subsidiar propostas de ações de saúde nos municípios sob sua responsabilidade.
3

Protocole de diagnostic des entraînements asynchrones par références : application à la détection des déséquilibres mécaniques et des défauts de courroies / Diagnosis protocol of induction machines by references : application to the detection of mechanical unbalances and belts faults

Fournier, Etienne 26 November 2015 (has links)
Les applications entrainées par des moteurs asynchrones consomment à elles seules la majorité de l’énergie électrique utilisée dans le secteur industriel. Des défauts, majoritairement d’origine mécanique, sont susceptibles de se développer au cours du fonctionnement de ces systèmes et rendent nécessaires la mise en place d’une maintenance souvent onéreuse. Afin de réduire ces coûts et d’optimiser la disponibilité des matériels, des méthodes de diagnostic peuvent être développées afin de détecter au plus tôt l’apparition d’un défaut. Cependant, la fiabilité de ces algorithmes et l’automatisation du diagnostic sont complexifiées par la diversité des systèmes alimentés par les variateurs de vitesse industriels ainsi que par le changement de leur point de fonctionnement. Afin de répondre à ces problématiques, un protocole de diagnostic permettant de réaliser une détection sensible et robuste des défaillances mécaniques est développé dans ce travail. Cette méthode consiste à segmenter le plan couple-vitesse du moteur surveillé et à réaliser des références statistiques caractérisant son fonctionnement sain sur chacune des zones de fonctionnement. Des indicateurs de défaut peuvent ensuite être définis statistiquement afin de quantifier l’écart du système vis-à-vis de son comportement sain. Ces indicateurs sont ainsi indépendants du système surveillé et de son point de fonctionnement. Les principales applications ciblées par ce protocole sont les systèmes de compression, de pompage et de ventilation qui représentent une part importante de l’énergie électrique consommée dans le secteur industriel. Le protocole de diagnostic développé dans cette étude est validé expérimentalement dans le cadre de la détection de différentes niveaux de déséquilibre mécanique mais également pour la surveillance d’un système de transmission de type poulies-courroies. De plus, les campagnes expérimentales sont effectuées sur plusieurs systèmes électromécaniques dont les propriétés sont foncièrement différentes afin de valider les performances de normalisation de ce protocole. / Systems driven by induction motors consume most of the industrial electric supply. Mechanical faults may develop over time and disrupt their proper functioning, resulting in high maintenance expenditure. In order to reduce these costs and to improve machines availability, diagnosis methods can be developped to perform the early detection of incipient faults. However, the variety of industrial applications driven by induction motors complicates the automation of fault detection. To resolve these issues, a sensitive and robust diagnosis protocol has been developed in two stages. First, the torque-speed plane of the monitored motor is segmented in different functioning areas. Then, a statistical reference is achieved on each operating zone in order to characterize the healthy working of the system. Statistic-based indicators may then quantify the system deviation from its healthy behaviour. These indicators are therefore independent of the monitored system and of its functioning point. The applications targeted by this study are fans, pumps and compressors which are widely used in industrial processes. This diagnosis protocol is experimentally validated on the detection of several levels of mechanical unbalance and for the monitoring of V-belts drive health state. Moreover, test campaigns are achieved on varied electromechanical systems in order to validate the normalization performances of the developed protocol.

Page generated in 0.1207 seconds