Critics argue that there has been a trend among Microfinance Institutions (MFI) to focus on profitability in order to stay financially sustainable. This made some institutions neglect the social mission of microfinancing. In this paper I intend to examine if empirical evidence supports this so called mission drift hypothesis as well as other claims in this context. Using the global panel data set of the MIX (Microfinance Information Exchange), which gathers from 1995 to 2010 and contains up to 1400 institutions with a high variety of organizational forms, I was able to identify a world-wide mission drift effect in their social goal of reaching out the poorest part of the population. Furthermore, I find that, on average, the outreach of an MFI has a significant negative influence on its short and long term financial performance. Despite that, I eventually proved that the probability that an MFI worsens its social performance substantially increases if its profitability has decreased in the previous years. / Das Konzept der Mikrofinanzierung wurde, insbesondere im Zuge der Mikrofinanzkrisen in Asien und Südamerika zunehmend kritisiert. Dabei stand vor allem die Kommerzialisierung der Branche im Zentrum der Kritik. In dieser Studie soll daher unter anderem die sogenannte „Mission Drifts”-These also dass das eigentliche Ziel des Mikrokreditwesen aus den Augen verloren wurde, empirisch überprüft werden. Mit Hilfe des Microfinance Information Exchange (MIX) Datensatzes, wurden Paneldaten von bis zu 1.400 Kreditinstitutionen, mit unterschiedlichen (Rechts-)formen, aus den Jahren 1995 bis 2010 ausgewertet. Die Regressionsanalyse hat gezeigt, dass Profitablität in der Tat einen negativen Einfluss auf das Ziel hat, möglichst arme Menschen zu erreichen. Auch der Trade-off zwischen der Reichweite von Mikrokrediten und kurzfristiger sowie langfristiger Profitabilität konnte nachgewiesen werden. Die Daten zeigten aber auch, dass Mikrofinanzinstitution dazu tendieren soziale Ziele zu vernachlässigen, wenn es im vergangenen Geschäftsjahr finanziell bergab ging.

The objective of our research is to develop system architectures and CMOS circuits that interface with high-Q silicon microgyroscopes to implement navigation-grade angular rate sensors. The MEMS sensor used in this work is an in-plane bulk-micromachined mode-matched tuning fork gyroscope (M² – TFG ), fabricated on silicon-on-insulator substrate. The use of CMOS transimpedance amplifiers (TIA) as front-ends in high-Q MEMS resonant sensors is explored. A T-network TIA is proposed as the front-end for resonant capacitive detection. The T-TIA provides on-chip transimpedance gains of 25MΩ, has a measured capacitive resolution of 0.02aF /√Hz at 15kHz, a dynamic range of 104dB in a bandwidth of 10Hz and consumes 400μW of power. A second contribution is the development of an automated scheme to adaptively bias the mechanical structure, such that the sensor is operated in the mode-matched condition. Mode-matching leverages the inherently high quality factors of the microgyroscope, resulting in significant improvement in the Brownian noise floor, electronic noise, sensitivity and bias drift of the microsensor. We developed a novel architecture that utilizes the often ignored residual quadrature error in a gyroscope to achieve and maintain perfect mode-matching (i.e.0Hz split between the drive and sense mode frequencies), as well as electronically control the sensor bandwidth. A CMOS implementation is developed that allows mode-matching of the drive and sense frequencies of a gyroscope at a fraction of the time taken by current state of-the-art techniques. Further, this mode-matching technique allows for maintaining a controlled separation between the drive and sense resonant frequencies, providing a means of increasing sensor bandwidth and dynamic range. The mode-matching CMOS IC, implemented in a 0.5μm 2P3M process, and control algorithm have been interfaced with a 60μm thick M2−TFG to implement an angular rate sensor with bias drift as low as 0.1°/hr ℃ the lowest recorded to date for a silicon MEMS gyro.

Ce travail de thèse se concentre sur l'étude des modèles de type drift-diffusion. Des approches sont développées pour la modélisation de la Microscopie à sonde de Kelvin, des cellules solaires à base de matériaux pérovskites (PSCs), des cellules solaires tandem de type pérovskite/silicium et des îlots quantiques lll-V/GaP. Tout d'abord, l'approche de la modélisation de la sonde de Kelvin est examinée pour la surface de TiOx et l'absorbeur pérovskite MAPbI3 Ensuite, des mesures avec sonde de Kelvin et des simulations sont proposées pour des jonctions diffuses à base de silicium et pour des PSCs à base de TiOx mésa poreux. Les variations du potentiel interne sont étudiées ouvrant la voie à une amélioration supplémentaire des dispositifs. L'influence de l'état de surface des couches wo. sur des mesures à sonde de Kelvin est étudiée théoriquement. Différents facteurs à l'origine des pertes de tension de circuit ouvert (Voc) des PSCs sont discutés. L'effet anormal d'hystérésis dans les PSCs est également simulé, en tenant compte des étals de pièges d'interface et des ions mobiles. En outre, le design de cellules solaires tandem 2T pérovskite/silicium est étudié en détails. Une jonction tunnel à base de silicium entre les deux sous-cellules supérieure et inférieure est proposée pour assurer le bon fonctionnement des cellules en série. L'influence du profil de dopage dans la jonction tunnel est discutée. Au final, le transport des porteurs dans les îlots quantiques III-V/GaP est étudié dans le cadre plus général de l'intégration d'émetteurs lll-V sur silicium. Les caractéristiques électroluminescentes et électriques de ces structures sont simulées dans une approximation cylindrique. / This PhD work focuses on optoelectronic device simulations based on drift-diffusion models. Approaches are developed for the modelling of Kelvin Probe Force Microscopy (KPFM), perovskite-based solar cells (PSCs), perovskite/silicon tandem solar cells and lll-V/GaP quantum dots (ODs). Firstly, a new approach for the modelling of KPFM is applied to TiOx slabs and to the MAPbI3 perovskite absorber. Secondly, KPFM measurements and simulations are proposed for silicon-based diffused junctions and mesoporous TiOx based PSCs. The built-in potential is investigated, and this study paves the way toward fu rther device improvements. In addition, the influence of the surface of WO. slabs on KPFM measurements is studied theoretically. Various facto rs influencing open circuit voltage (Voe) losses in PSCs are discussed. The abnormal hysteresis effect in the PSCs is simulated as well, considering interface trap states and mobile ions. The design of two-terminal perovskite/silicon tandem solar cells is studied in detail. A siliconbased tunnel junction between the top and the bottom subcells is proposed for serial current matching. The influence of the doping profile in the tunnel junction is discussed. At the end of the manuscript, the carrier transport in III-V/GaP QDs is investigated, for the integration of III-V emitters on silicon. The electroluminescence and electrical characteristics of these III -V light emitting devices are simulated by using a cylindrical approximation.

O desempenho das Máquinas de Medir a Três Coordenadas (MM3Cs) fica limitado por diversos fatores, que atuam de maneira conjunta gerando os denominados erros volumétricos. Para a temperatura de 20ºC os erros geométricos podem ser considerados constantes, uma vez que variam muito lentamente com o tempo. Porém, se a temperatura é alterada estes erros mudam em grandeza e comportamento, gerando os denominados erros térmicos. Alguns trabalhos têm sido desenvolvidos com o objetivo de estudar e modelar os erros térmicos, porém os resultados alcançados são, ainda, incipientes. Este trabalho apresenta o equacionamento das componentes do erro volumétrico das MM3Cs considerando as influências térmicas. A medelagem foi aplicada a uma MM3C do tipo \"Ponte Móvel\" e combina transformações homogêneas, técnicas de regressão e mínimos quadrados. As grandezas dos erros geométricos e das variações termicamente induzidas destes erros foram coletadas utilizando-se do interferômetro laser, do esquadro mecânico, do nível eletrônico, etc. Os valores das temperaturas foram monitorados através de termopares do tipo T (Cobre-Constantan). Verificou-se que a Máquina não experimenta deformações, além, das provocadas pela livre dilatação dos seus componentes. A partir do modelo proposto foram sintetizadas as componentes do erro volumétrico, os resultados foram discutidos e comparados com aqueles obtidos através da medição de um anel padrão, constatando-se a excelente capacidade do modelo na previsão do erro volumétrico da máquina. No caso, erros da ordem de grandeza de 10 &#956m foram reduzidos em pelo menos 75%, enquanto que para erros maiores que 10 &#956m a eficiência do modelo foi de 90%. / Performance of coordinate measuring machines (CMMs) is limited by numerous factors that operate simultaneously and generate volumetric errors. The most significant portion of the volumetric error is produced by geometric errors. At the temperature of 20ºC, geometric errors can be considered at steady states, once their variation in time is considerably slow. However, if temperature is modified, these errors change in magnitude and behaviour, generating the thermal induced errors. Some work has been developed aiming to study and model the thermal errors, but the achieved results are still incipient. This work presents the derivation of the components of the volumetric error considering its thermal influences. The method was employed and applied to moving bridge CMM and combines homogeneous transformation, regression techniques and least squares methods. The magnitudes of the geometric errors and its thermally induced variations were collected by means of a laser interferometer system, mechanical square, electronic level, etc. Temperature data were monitored by means of T-type thermocouples (copper-constantan). It was verified that the CMM was not susceptible to deformations other than the ones due to the dilatation of its components. From the proposed model, the components of volumetric error were synthesized; the results were discussed and compared to the ones obtained from the measurement of a ring plug, observing the outstanding ability of the model to predict the volumetric error of the machine. Errors of 10 &#956m in magnitude were reduced in at least 75%, whilst errors greater than 10 &#956m, presented a reduction efficiency of 90%. It was verified that the CMM was not susceptible to deformations other than the ones due to the dilatation of its components.

A pandemic subtype of influenza A sometimes replaces (e.g., in 1918, 1957, 1968) but sometimes coexists (e.g., in 1977) with the previous seasonal subtype. This research aims to determine a condition for replacement or coexistence of influenza subtypes. We formulate a hybrid model for the dynamics of influenza A epidemics taking into account cross-immunity of influenza strains depending on the most recent seasonal infection. A combination of theoretical and numerical analyses shows that for very strong cross-immunity between seasonal and pandemic subtypes, the pandemic cannot invade, whereas for strong and weak cross-immunity there is coexistence, and for intermediate levels of cross-immunity the pandemic may replace the seasonal subtype. Cross-immunity between seasonal strains is also a key factor of our model because it has a major influence on the final size of seasonal epidemics, and on the distribution of susceptibility in the population. To determine this cross-immunity, we design a novel statistical method, which uses a theoretical model and clinical data on attack rates and vaccine efficacy among school children for two seasons after the 1968 A/H3N2 pandemic. This model incorporates the distribution of susceptibility and the dependence of cross-immunity on the antigenic distance of drifted strains. We find that the cross-immunity between an influenza strain and the mutant that causes the next epidemic is 88%. Our method also gives an estimated value 2.15 for the basic reproduction number of the 1968 pandemic influenza. Our hybrid model agrees qualitatively with the observed subtype replacement or coexistence in 1957, 1968 and 1977. However, our model with the homogeneous mixing assumption significantly over estimates the pandemic attack rate. Thus, we modify the model to incorporate heterogeneity in the contact rate of individuals. Using the determined values of cross-immunity and the basic reproduction number, this modification lowers the pandemic attack rate slightly, but it is still higher than the observed attack rates. / Graduate

This study focuses on non-ductile reinforced concrete (RC) frames built outside current practices. These structures are quite vulnerable to collapse during earthquakes. One option to retrofit buildings with poorly detailed RC columns is to construct full-height masonry infill walls to provide additional means to resist loads caused by gravity and increase lateral stiffness resulting in a reduction in drift demand. On the other hand, infill can cause reductions in drift capacity that offset the benefits of reductions in drift demand. Given these two opposing effects, this investigation addresses the following question: are poorly detailed RC frames with masonry infill walls any safer than similar RC frames without infill walls?

Stockholms spårvägar växer i allt snabbare fart, och det är allt fler kommuner som är intresserade av den typen av transportmedel. Något som idag är en brist inom spårvägsindustrin i Stockholm är kunskap och erfarenhet om spårvägsämnet. Drift och underhåll har varit minimal eftersom det inte har prioriterats sedan Tvärbanan öppnades. Detta har lett till för tidig reinvestering av gatuspår i mångmiljonbelopp. Då det förekommer olika förutsättningar som måste tas hänsyn till i form av buller, mån om plats och geoteknik finns det många konstruktioner att välja mellan när en spårväg ska byggas. Denna rapport har som syfte att underlätta val av spårvägsinvesteringar i framtiden. Det har också som mål att bidra med kunskap om hur viktigt det är med drift- och underhåll och att det inte bör snålas in på den aspekten. En begränsning har gjorts till fem spårvägskonstruktioner som förekommer i Stockholm och vanligtvis i Göteborg och Norrköping. För att få en förståelse för hur konstruktionerna är uppbyggda redogörs en beskrivning av dem i rapporten.  Rapporten inleds med hur det historiskt har sett ut med spårvägar, nedläggning av dem och återinvestering i Stockholm. Dessa år av glapp mellan nedläggning och återinvestering i Stockholm visar på hur mycket kunskap som har glömts bort och som måste fås tillbaka för framtida investeringar. En analys kring placering av de olika konstruktionerna, hur varierande för- och nackdelarna och kostnaderna är avslutar rapporten. Mycket av rapporten bygger på litteraturstudier på tidigare rapporter som har kompletterats med kontinuerliga möten med handledare och kontakt med personer från olika kommuner och företag i Sverige för att inhämta trovärdig information. / The tramways of Stockholm are growing at an increasing rate and so is the number of municipals interested in that type of transportation.  In Stockholm’s tramway industry there is currently a lack of knowledge and experience on the subject of tramways. Drift and maintenance has been minimal due to lack of prioritization since Tvärbanan was opened. This has led to a premature reinvestment of millions of SEK for road tracks. Due to different presumptions that have to be taken into account such as noise, considerations of room and geoscience, there are many different constructions to choose from when a tramway is to be built. This report’s purpose serves to facilitate choice of tramway investments in the future. Another goal of the report is to contribute with knowledge of how important service and maintenance of tramways is and why shying away from spending money on these matters is bad. For the purpose of this report a restriction has been made, five tramway constructions will be looked at, these are present in Stockholm but also in Göteborg and Norrköping. To acquire an understanding of how the constructions are built, a description of these is accounted for in the report. The report starts off with a historical perspective on tramways, closure of them and reinvestment of them in Stockholm. The years of gaps between closure and reinvestment in Stockholm shows how much knowledge has been forgotten and has to be reacquired for future investments. In the end of the report an analysis, about placement of the different constructions, how varied the pros and cons are and the costs of these, is presented. Much of the report is based on literature studies of earlier reports and has been complemented with continuous meetings with tutors and contact with people from different municipals and companies in Sweden to obtain trustworthy information.

Regression models are often required for controlling production processes by predicting parameter values. However, the implicit assumption of standard regression techniques that the data set used for parameter estimation comes from a stationary joint distribution may not hold in this context because manufacturing processes are subject to physical changes like wear and aging, denoted as process drift. This can cause the estimated model to deviate significantly from the current state of the modeled system. In this paper, we discuss the problem of estimating regression models from drifting processes and we present ensemble regression, an approach that maintains a set of regression models—estimated from different ranges of the data set—according to their predictive performance. We extensively evaluate our approach on synthetic and real-world data.

This dissertation integrates a process-based hydrological investigation with an ongoing paired-catchment study to better understand how forest harvest impacts catchment function at multiple scales. We do this by addressing fundamental questions related to the stocks, flows and transit times of water. Isotope tracers are used within a top-down catchment intercomparison framework to investigate the role of geology in controlling streamwater mean transit time and their scaling relationships with the surrounding landscape. We found that streams draining catchments with permeable bedrock geology at the Drift Creek watershed in the Oregon Coast Range had longer mean transit times than catchments with poorly permeable bedrock at the HJ Andrews Experimental Forest in the Oregon Cascades. We also found that differences in permeability contrasts within the subsurface controlled whether mean transit time scaled with indices of catchment topography (for the poorly permeable bedrock) or with catchment area (for the permeable bedrock). We then investigated the process-reasons for the observed differences in mean transit time ranges and scaling behavior using a detailed, bottom-up approach to characterize subsurface water stores and fluxes. We found that the mean transit times in catchments underlain by permeable bedrock were influenced by multiple subsurface storage pools with different groundwater ages, whereas storage in the poorly permeable catchments was limited to the soil profile and that resulted in quick routing of excess water to the stream at the soil bedrock interface, leading to mean transit times that were closely related to flowpath lengths and gradients. Finally, we examined how and where forest trees interacted with subsurface storage during the growing season using a forest manipulation experiment, where we tested the null hypothesis that near-stream trees alone influenced daily fluctuations in streamflow. We felled trees within this zone for two 2.5 ha basins and combined this with isotopic tracing of tree xylem water to test if water sources utilized by trees actively contributed to summer streamflow. We rejected our null hypotheses and found that diel fluctuations in streamflow were not generated exclusively in the near-stream zone. We were unable to link, isotopically, the water sources trees were utilizing to water that was contributing to streamflow. Our results provide new process-insights to how water is stored, extracted, and discharged from our forested catchments in Western Oregon that will help better explain how forest removal influences streamflow across multiple scales and geological conditions. / Graduation date: 2012

Submitted by Fabio Sobreira Campos da Costa (fabio.sobreira@ufpe.br) on 2016-07-11T12:33:28Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) silas-dissertacao-versao-final-2016.pdf: 1708159 bytes, checksum: 6c0efc5f2f0b27c79306418c9de516f1 (MD5) / Made available in DSpace on 2016-07-11T12:33:28Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) silas-dissertacao-versao-final-2016.pdf: 1708159 bytes, checksum: 6c0efc5f2f0b27c79306418c9de516f1 (MD5) Previous issue date: 2015-02-27 / FACEPE / A extração de conhecimento em ambientes com fluxo contínuo de dados é uma atividade que vem crescendo progressivamente. Diversas são as situações que necessitam desse mecanismo, como o monitoramento do histórico de compras de clientes; a detecção de presença por meio de sensores; ou o monitoramento da temperatura da água. Desta maneira, os algoritmos utilizados para esse fim devem ser atualizados constantemente, buscando adaptar-se às novas instâncias e levando em consideração as restrições computacionais. Quando se trabalha em ambientes com fluxo contínuo de dados, em geral não é recomendável supor que sua distribuição permanecerá estacionária. Diversas mudanças podem ocorrer ao longo do tempo, desencadeando uma situação geralmente conhecida como mudança de conceito (concept drift). Neste trabalho foi realizado um estudo comparativo entre alguns dos principais métodos de detecção de mudanças: ADWIN, DDM, DOF, ECDD, EDDM, PL e STEPD. Para execução dos experimentos foram utilizadas bases artificiais – simulando mudanças abruptas, graduais rápidas, e graduais lentas – e também bases com problemas reais. Os resultados foram analisados baseando-se na precisão, tempo de execução, uso de memória, tempo médio de detecção das mudanças, e quantidade de falsos positivos e negativos. Já os parâmetros dos métodos foram definidos utilizando uma versão adaptada de um algoritmo genético. De acordo com os resultados do teste de Friedman juntamente com Nemenyi, em termos de precisão, DDM se mostrou o método mais eficiente com as bases utilizadas, sendo estatisticamente superior ao DOF e ECDD. Já EDDM foi o método mais rápido e também o mais econômico no uso da memória, sendo superior ao DOF, ECDD, PL e STEPD, em ambos os casos. Conclui-se então que métodos mais sensíveis às detecções de mudanças, e consequentemente mais propensos a alarmes falsos, obtêm melhores resultados quando comparados a métodos menos sensíveis e menos suscetíveis a alarmes falsos. / Knowledge extraction from data streams is an activity that has been progressively receiving an increased demand. Examples of such applications include monitoring purchase history of customers, movement data from sensors, or water temperatures. Thus, algorithms used for this purpose must be constantly updated, trying to adapt to new instances and taking into account computational constraints. When working in environments with a continuous flow of data, there is no guarantee that the distribution of the data will remain stationary. On the contrary, several changes may occur over time, triggering situations commonly known as concept drift. In this work we present a comparative study of some of the main drift detection methods: ADWIN, DDM, DOF, ECDD, EDDM, PL and STEPD. For the execution of the experiments, artificial datasets were used – simulating abrupt, fast gradual, and slow gradual changes – and also datasets with real problems. The results were analyzed based on the accuracy, runtime, memory usage, average time to change detection, and number of false positives and negatives. The parameters of methods were defined using an adapted version of a genetic algorithm. According to the Friedman test with Nemenyi results, in terms of accuracy, DDM was the most efficient method with the datasets used, and statistically superior to DOF and ECDD. EDDM was the fastest method and also the most economical in memory usage, being statistically superior to DOF, ECDD, PL and STEPD, in both cases. It was concluded that more sensitive change detection methods, and therefore more prone to false alarms, achieve better results when compared to less sensitive and less susceptible to false alarms methods.