Parameter estimation is a powerful and adaptable framework that addresses the inherent complexities and uncertainties of financial data. We provide an overview of likelihood functions, and likelihood estimations, as well as the essential numerical approximations and techniques. In the financial domain, where unpredictable and non-stationary market dynamics prevail, parameter estimations of relevant SDE models prove highly relevant. We delve into practical applications, showcasing how SDEs can effectively capture the inherent uncertainties and dynamics of financial models related to time evolution of interest rates. We work with the Vašíček model and Cox-Ingersoll-Ross (CIR) model which describes the dynamics of interest rates over time. We incorporate the Maximum likelihood and Quasi-maximum likelihood estimation methods in estimating the parameters of our models.

HR-funktionen har genomgått en del förändringar under de senaste åren. Förändringarna har främst skett genom att HR har gått från att vara en personaladministrativ funktion till en allt mer strategisk funktion. Detta har ställt nya krav på HR-funktionens roll och dess arbete. Bland annat har HR behövt bli mer datadrivna. Störst utmaning har detta inneburit för HR-funktioner inom den offentliga sektorn, som till skillnad från privat sektor, fortfarande ligger efter i den datadrivna utvecklingen, och lite forskning har gjorts i den offentliga kontexten. Syftet med denna masteruppsats är därför att öka förståelsen för datadrivet HR-arbete inom offentlig sektor. För att göra detta undersöktes tre forskningsfrågor: (1) Hur påverkas HR-funktionens roll av datadrivet HR-arbete? (2) Vilka möjligheter och utmaningar finns med datadrivet HR-arbete? (3) Vad är unikt med datadriven HR i offentlig sektor? För att få svar på forskningsfrågorna har en kvalitativ flerfallsstudie genomförts i nio olika kommuner i Västra Götaland och Halland. Data samlades in genom semistrukturerade intervjuer och en tematisk analysmetod tillämpades vid sammanställningen av resultatet. Resultatet visar att HR-funktionens roll och deras arbete drivs mer i linje med rationell styrfilosofi när ett mer datadrivet HR-arbete tillämpas. Denna studie visar att datadriven HR riskerar att ta bort HR-funktionens fokus på mjuk HR, och istället främja hård HR inom HR-arbetet, där mätning och kontroller förekommer i större grad. Möjligheten med datadrivet HR-arbete är att kunna använda och nyttja stora mängder tillgänglig data för att ta bättre beslut och genomföra effektiva insatser inom HR-området. Det möjliggör ett mer strategiskt HR-arbete som kan bidra till ökad legitimitet och förtroende för HR-funktionen. Utmaningen är att det ofta saknas kompetens hos HR-personal att arbeta datadrivet. Resultatet visar också att offentlig sektor påverkas av det faktum att de förväntas tillämpa en allt mer tillitsbaserad styrning, vilket är en styrfilosofi som kan anses vara oförenlig med datadriven HR. Studien bidrar med djupare förståelse, och fyller en kunskapslucka, gällande datadrivet HR-arbete inom offentlig sektor. Studien bidrar med insikter gällande den stora avvägning HR-funktionen är på väg in i, när HR måste hitta sätt att integrera både mjuk och hård. HR-funktionerna står mitt emellan en betydande möjlighet av att använda data, samtidigt som det finns en kraft i tillitsbaserad styrning som menar att offentlig sektor måste sluta att mäta allting, då det kan skapa ett alltför stort fokus på kontroll och detaljstyrning. Utmaningen med datadriven HR är därför hur HR-funktionerna ska använda stora mängder data utan att bli allt för datadrivna. Studien belyser att det i datadriven HR måste finnas kunskap kring när mätningar kan anses vara säkra och användas för att göra förutsägelser eller utvärderingar, men också när mätningar kan skapa felaktiga incitament eller distraktioner. Denna studies resultat kan användas som ett hjälpmedel för HR-funktioner inom den offentliga sektorn och fungera som ett ramverk för dessa HR-funktioner i syfte att utvärdera sitt datadrivna HR-arbete.

<div>The app-based taxi services (ATS) has disrupted the traditional (street-hailing) taxi services (TTS) leading to transformative changes in the urban taxi markets and its impacts on mobility, design and environment. However, the current modeling of these new mobility markets is limited in its understanding of: (1) the underlying factors that influence the growth of the ATS market; (2) the competition of ATS and TTS markets; (3) pricing in the ATS market; (4) system wide tools to understand the impacts of the market. The overarching goal of this dissertation is to address four fundamental processes of taxi system, ranging from demand generation, supply generation and exiting, dynamic pricing generation, and vehicle-passenger matching over road network. This dissertation achieves these goals by using original large scale datasets to characterize disruptive changes in mobility, understand strategic behaviors of stakeholders, and formulate system dynamics.</div><div> </div><div>This dissertation develops various modeling structures and estimation methods, motivated from statistical, econometric, machine learning, and stochastic approaches. First, we adapt multiple econometric models for demand, supply, and platform-exiting (offline) behaviors, including mixture model of spatial lag and Poisson regression and mixture model of spatial lag and panel regression. It is apparent that all proposed econometric models should be corrected with spatial lag due to significant spatial autocorrelations. The results indicate effectiveness of dynamic pricing in controlling demand, however, it also shows no impacts on driver's online and offline behaviors. Then a dynamic pricing generation problem is formulated with multi-class classification. This model is empirically validated for the impacts of demand and supply in dynamic price generation and the significant spatial and temporal heterogeneity. Last, we propose a queueing network consisting of taxi service queues for vehicle-passenger matching and road service queue for vehicle movements at homogeneous spatial units. The method captures stochasticity in vehicle-passenger matching process, and more importantly, formulates the interactions with urban road traffic.</div><div> </div><div>In summary, this dissertation provides a holistic understanding of fundamental processes that govern the rapid rise in ATS markets and in developing quantitative tools for the system wide impacts of this evolving taxi markets. Taken together, these tools are transformative and useful for city agencies to make various decisions in the smart mobility landscape. </div>

Electrical contacts provide a well-proven solution to switching various loads in a wide variety of applications, such as power distribution, control applications, automotive and telecommunications. However, electrical contacts are known for limited reliability due to degradation effects upon the switching contacts due to arcing and fretting. Essentially, the life of the device may be determined by the limited life of the contacts. Failure to trip, spurious tripping and contact welding can, in critical applications such as control systems for avionics and nuclear power application, cause significant costs due to downtime, as well as safety implications. Prognostics provides a way to assess the remaining useful life (RUL) of a component based on its current state of health and its anticipated future usage and operating conditions. In this thesis, the effects of contact wear on a set of electromagnetic relays used in an avionic power controller is examined, and how contact resistance combined with a prognostic approach, can be used to ascertain the RUL of the device. Two methodologies are presented, firstly a Physics based Model (PbM) of the degradation using the predicted material loss due to arc damage. Secondly a computationally efficient technique using posterior degradation data to form a state space model in real time via a Sliding Window Recursive Least Squares (SWRLS) algorithm. Health monitoring using the presented techniques can provide knowledge of impending failure in high reliability applications where the risks associated with loss-of-functionality are too high to endure. The future states of the systems has been estimated based on a Particle and Kalman-filter projection of the models via a Bayesian framework. Performance of the prognostication health management algorithm during the contacts life has been quantified using performance evaluation metrics. Model predictions have been correlated with experimental data. Prognostic metrics including Prognostic Horizon (PH), alpha-Lamda (α-λ), and Relative Accuracy have been used to assess the performance of the damage proxies and a comparison of the two models made.

abstract: The purpose of this action research was to understand how reflective, job-embedded early childhood science professional learning and development (PLD) impacted Early Head Start (EHS) teacher learning and their perceptions toward science with toddlers. Limited content knowledge and lack of formal preparation impact teachers’ understanding of developmentally appropriate science and their capacity to support children to develop science skills. In Arizona, limited availability of early childhood science coursework and no science-related PLD for toddler teachers showed the need for this project. Four literature themes were reviewed: teacher as researcher, how people learn, reflective PLD, and how young children develop scientific thinking skills. The participants were nine EHS teachers who worked at the same Head Start program in five different classrooms in Arizona. The innovation included early childhood science workshops, collaboration and reflecting meetings (CPRM), and electronic correspondence. These were job-embedded, meaning they related to the teachers’ day-to-day work with toddlers. Qualitative data were collected through CPRM transcripts, pre/post-project interviews, and researcher journal entries. Data were analyzed using constant comparative method and grounded theory through open, focused, and selective coding. Results showed that teachers learned about their pedagogy and the capacities of toddlers in their classrooms. Through reflective PLD meetings, teachers developed an understanding of toddlers’ abilities to engage with science. Teachers acquired and implemented teacher research skills and utilized the study of documentation to better understand children’s interests and abilities. They recognized the role of the teacher to provide open-ended materials and time. Moreover, teachers improved their comfort with science and enhanced their observational skills. The teachers then saw their role in supporting science as more active. The researcher concluded that the project helped address the problem of practice. Future research should consider job-embedded PLD as an important approach to supporting data-driven instructional practices and reflection about children’s capabilities and competencies. Keywords: action research, Arizona Early Childhood Workforce Knowledge and Competencies, Arizona’s Infant and Toddler Developmental Guidelines (ITDG), documentation, early childhood science, Early Head Start (EHS), Head Start Early Learning Outcomes Framework (ELOF), inquiry, job-embedded, pedagogy, professional development (PD), reflective professional development, teacher as researcher, teacher research, toddler science / Dissertation/Thesis / Doctoral Dissertation Leadership and Innovation 2019

School leaders at an urban high school in the U.S. Midwest encouraged teachers to use formative assessment to help students meet learning goals; however, several years later, they found inconsistent implementation. Without a clear understanding of teachers' formative assessment practices, leaders could not establish needed supports for its consistent use in the classrooms. The purpose of this bounded qualitative case study was to examine teachers' formative assessment use to check for student understanding and to adjust instruction. Black and Wiliam's formative assessment theory formed the foundation of this study. Research questions focused on teachers' perceptions of formative assessment and usage of formative assessment for instruction. Ten state certified high school teachers, who had at least a bachelor's degree, passed basic skills and subject area examinations, and taught within their majors or minors, were purposefully selected to provide data. Data were gathered from observations, interviews, and teacher logs and were analyzed inductively using open and axial coding strategies. Results showed teachers collected and used formative assessment to modify instruction and determine student understanding from a limited number of students. Furthermore, they lacked the knowledge, skills, and strategies to implement formative assessment to help all students meet learning goals. Based on the findings, 3 professional development (PD) sessions were created to help school leaders provide support for teachers' consistent formative assessment implementation. These endeavors may contribute to positive social change when administrators provide teachers with PD to increase teachers' knowledge and skills using formative assessment, and, ultimately, to meet student learning goals.

The strategic prevention framework (SPF) is a data-driven operating system to assist designing evidence-based substance abuse prevention programs. The study performed here was to assess the effectiveness of the SPF as a prevention planning system. One purpose of this study was to determine the implementation fidelity of the programs that used the SPF process; the other purpose was to assess effectiveness of the SPF process. This study utilized a set of data collected by the national cross site evaluation team on all jurisdictions that implemented the SPF. A subset of communities collected and reported at least 2 pre-implementation and at least 2 post implementation outcomes data. The minimum sample size for the study was determined by using Cohen's d criteria. The assessments were performed using both qualitative and quantitative methods by using data collected from multiple levels with a quasi-experimental design. The qualitative data were analyzed using qualitative software with key word searches to examine implementation processes, and the quantitative data were analyzed using descriptive statistics and inferential methods such as Student t tests to examine and compare outcomes. Results show that the communities in the study implemented the SPF process with fidelity and that there were changes in desired directions. Factors related to improvements include sufficient internal resources and monitoring follow-through. This research has important implications for social change since substance abuse is a major social issue that has consequences across life span. Recent studies have shown that many behavioral problems have similar risk factors and that improvements for some behavioral problems will most likely have beneficial effects on other related problems.

<p>Citizens’ interaction with governments is an area with unique implications for channel management. Governments need to take the citizens perspective into further consideration in order to be successful in delivering high-quality e-services. This paper aims to determine if a categorization of citizen-initiated contacts from a citizen-centric perspective can be a valuable basis for decisions regarding e-service development. The study consisted of three steps. The first step was an examination of the existing related literature, which resulted in the uncovering of the most important aspects of citizens channel choice. The second step consisted of an elaboration of a classification based on perceived task characteristics and a subsequent matching of the categories to desirable channel characteristics. The third and final step consisted of an application of the proposed categorization on a content management system containing all citizen-initiated contacts in a Swedish municipality. The application indicated that the proposed categorization could possibly be used to guide investments in e-services towards a channel-appropriate direction.</p>

Le cadre principal de cette thèse est la classification non supervisée, traitée par une approche statistique dans le cadre des modèles de mélange. Plus particulièrement, nous nous intéressons au choix du nombre de classes et au critère de sélection de modèle ICL. Une approche fructueuse de son étude théorique consiste à considérer un contraste adapté à la classification non supervisée : ce faisant, un nouvel estimateur ainsi que de nouveaux critères de sélection de modèle sont proposés et étudiés. Des solutions pratiques pour leur calcul s'accompagnent de retombées positives pour le calcul du maximum de vraisemblance dans les modèles de mélange. La méthode de l'heuristique de pente est appliquée pour la calibration des critères pénalisés considérés. Aussi les bases théoriques en sont-elles rappelées en détails, et deux approches pour son application sont étudiées. Une autre approche de la classification non supervisée est considérée : chaque classe peut être modélisée elle-même par un mélange. Une méthode est proposée pour répondre notamment à la question du choix des composantes à regrouper. Enfin, un critère est proposé pour permettre de lier le choix du nombre de composantes, lorsqu'il est identifié au nombre de classes, à une éventuelle classification externe connue a priori.

Achieving the Dream is a national initiative focused on helping more community college students succeed, particularly students of color and low-income students. Achieving the Dream’s student-centered model of institutional improvement focuses on eliminating gaps and raising student achievement by helping institutions build a culture of evidence through institutional transformation. This interpretive case study employed a mixed methods approach and utilized a sequential explanatory strategy to gather detailed information related to the research questions. The study examined, from an insider’s perspective, the progress made by an Achieving the Dream college in implementing practices that reflect the principles of the Achieving the Dream model of institutional improvement. The four principles of the model are: committed leadership, use of evidence to improve programs and services, broad engagement, and systemic institutional improvement. The study was conducted in two phases and involved a quantitative survey of all college employees and semi- structured, individual interviews with members of the college’s Achieving the Dream team. The quantitative and qualitative data were given equal weighting in the study and were integrated to the extent that the qualitative data collected provided supportive insights into the findings derived from the quantitative analysis. This study found the college made progress in implementing the practices reflected by the Achieving the Dream model. The study findings also provide insights into underlying factors that existed at the college related to its implementation of the initiative. The findings will supplement current knowledge about the efficacy of change models designed to help community colleges build a student-centered model of institutional improvement and the knowledge gained should not only benefit the college, but also provide useful information to other institutions involved in this initiative. / text