Educators across the country are expected to be data literate. They must be able to systemically collect and analyze student data to make informed instructional decisions. However, many school leaders lack the knowledge about how to transform mountains of data on student achievement into an action plan that will improve instruction and increase student learning (Boudett, et al., 2007). In addition, time constraints make it difficult for educators to effectively and efficiently collaborate around student data consistently. Most of the research on data use describes the importance for educators to use data to improve student achievement. However, limited research has been documented on the role the campus leader employs when creating a culture of data-driven decision-making as it relates to student achievement. Furthermore, the research on data use in Title 1 schools is also limited. Therefore, it is imperative to examine and describe how a Title 1 middle school principal implemented Data Teams on a campus. Consequently, the goal of this research was to determine how school leaders improve student learning through teacher data teams. The four primary questions this research addressed in this single case study were: 1. What is the role of the principal in implementing successful Data Teams? 2. What campus structures foster the Data Team process? 3. What are the perceptions of teachers regarding the effectiveness of the Data Team? 4. What practices contribute to the sustainability of Data Teams? Data was gathered through semi-structured interviews, direct observations, and document reviews which informed the findings. This research study revealed that the principal played a key leadership role in creating a culture of collaboration and data inquiry by implementing teacher Data Teams. Such leadership role is enacted by: communicating a vision for Data Teams, providing for job-embedded professional development, and offering differentiated support. Structured time, structured meetings, student data system, and structured assessments are structures employed by the school. Student-focused collaboration, enhanced teacher trust, and increased student achievement illustrate evidence of Data Team effectiveness. Shared accountability, building school culture, and focused interventions serve to sustain Data Teams. In conclusion, it can be affirmed the principal has the most influence on what will be supported on a campus. Therefore, the leadership role performed by the principal when guiding a faculty through the implementation of Data Teams must be deliberate and thoughtful. The principal should include key stakeholders in the decision-making process and build capacity among teachers to ensure the sustainability of Data Teams. Furthermore, targeted professional development and structures that allow time for teachers to collaborate are necessary. Because the ultimate goal for schools is student learning, it is important that everyone within the school organization understand their role in the Data Team process. / text

The Capacitance Resistance Model (CRM) is a reservoir model based on a data-driven approach. It stems from the continuity equation and takes advantage of the usually abundant rate data to achieve a synergy of analytical model and data-driven approach. Minimal information (rates and bottom-hole pressure) is required to inexpensively characterize the reservoir. Important information, such as inter-well connectivity, reservoir compressibility effects, etc., can be easily and readily evaluated. The model also suggests optimal injection schemes in an effort to maximize ultimate oil recovery, and hence can assist real time reservoir analysis to make more informed management decisions. Nevertheless, an important limitation in the current CRM model is that it only treats the reservoir flow as single-phase flow, which does not favor capturing physics when the saturation change is large, such as for an immature water flood. To overcome this limitation, we develop a two-phase flow coupled CRM model that couples the pressure equation (fluid continuity equation) and the saturation equation (oil mass balance). Through this coupling, the model parameters such as the connectivity, the time constant, temporal oil saturation, etc., are estimated using nonlinear multivariate regression to history match historical production data. Incorporating the physics of two-phase displacement brings several advantages and benefits to the CRM model, such as the estimation of total mobility change, more accurate prediction of oil production, broader model application range, and better adaptability to complicated field scenarios. Also, the estimated saturation within the drainage volume of each producer can provide insights with respect to the field remaining oil saturation distribution. Synthetic field case studies are carried out to demonstrate the different capabilities of the coupled CRM model in homogeneous and heterogeneous reservoirs with different geological features. The physical meanings of model parameters are well explained and validated through case studies. The results validate the coupled CRM model and show improved accuracy in model parameters obtained through the history match. The prediction of oil production is also significantly improved compared to the current CRM model. A more reliable oil rate prediction enables further optimization to adjust injection strategies. The coupled CRM model has been shown to be fast and stable. Moreover, sensitivity analyses are conducted to study and understand the impact of the input information (e.g., relative permeability, viscosity) upon the output model parameters (e.g., connectivity, time constants). This analysis also proves that the model parameters from the two-phase coupled model can combine both reservoir compressibility and mobility effects. / text

This study examined how higher performing middle schools in the Rio Grande Valley use data to drive instructional decisions. Three research questions guided this study: (a) to what extent do higher performing, Title-1, middle schools in the Rio Grande Valley utilize data to make schoolwide instructional decisions; (b) how does the principal support data use for instructional decision-making; and (c) what do teachers perceive to be the processes that have led to the current level of data use in instructional decision making? A mixed-methods multiple-case study included middle schools that were drawn from a list of higher performing schools according to Just for the Kids and the National Center for Educational Achievement. To be included in the study, schools had to be located in the Rio Grande Valley, Texas, specifically in the counties of Starr, Cameron or Hidalgo. Additionally, the schools needed to be designated a Title-1 school, according to federal criteria. Data for the study was collected using a survey, followed by one-on-one interviews. Descriptive analyses was then conducted using the survey data. The interview data was analyzed using first-level coding followed by the use of cross case analysis to determine themes common to all cases. The findings from this research revealed that data is used extensively in the schools studied; primarily to determine the instructional scope of what is taught. It was found that while data use was extensive, the source and purpose of data use was limited to that which was directly tied to the state-administered assessment (TAKS). The second major finding was that principals create the necessary conditions for data use that becomes an embedded practice, where teachers can take risks with their colleagues in reviewing and using data. This study concludes that more principals can lead their schools to greater levels of data use by creating the necessary conditions for change. At the same time, the findings suggest that there is a need for leaders at all levels to examine and mitigate the unintended consequences of data use that is derived from a single-source and for a single purpose—that is, performance on the state exam (TAKS). / text

In many applications, such as conversational agents, virtual reality, movies, and games, animated facial expressions of computer-generated (CG) characters are used to communicate, teach, or entertain. With an increased demand for CG characters, it is important to animate accurate, realistic facial expressions because human facial expressions communicate a wealth of information. However, realistically animating faces is challenging and time-consuming for two reasons. First, human observers are adept at detecting anomalies in realistic CG facial animations. Second, traditional animation techniques based on keyframing sometimes approximate the dynamics of facial expressions or require extensive artistic input while high-resolution performance capture techniques are cost prohibitive. In this thesis, we develop a framework to explore representations of two key facial expressions, blinks and smiles, and we show that data-driven models are needed to realistically animate these expressions. Our approach relies on utilizing high-resolution performance capture data to build models that can be used in traditional keyframing systems. First, we record large collections of high-resolution dynamic expressions through video and motion capture technology. Next, we build expression-specific models of the dynamic data properties of blinks and smiles. We explore variants of the model and assess whether viewers perceive the models as more natural than the simplified models present in the literature. In the first part of the thesis, we build a generative model of the characteristic dynamics of blinks: fast closing of the eyelids followed by a slow opening. Blinks have a characteristic profile with relatively little variation across instances or people. Our results demonstrate the need for an accurate model of eye blink dynamics rather than simple approximations, as viewers perceive the difference. In the second part of the thesis, we investigate how spatial and temporal linearities impact smile genuineness and build a model for genuine smiles. Our perceptual results indicate that a smile model needs to preserve temporal information. With this model, we synthesize perceptually genuine smiles that outperform traditional animation methods accompanied by plausible head motions. In the last part of the thesis, we investigate how blinks synchronize with the start and end of spontaneous smiles. Our analysis shows that eye blinks correlate with the end of the smile and occur before the lip corners stop moving downwards. We argue that the timing of blinks relative to smiles is useful in creating compelling facial expressions. Our work is directly applicable to current methods in animation. For example, we illustrate how our models can be used in the popular framework of blendshape animation to increase realism while keeping the system complexity low. Furthermore, our perceptual results can inform the design of realistic animation systems by highlighting common assumptions that over-simplify the dynamics of expressions.

No description available.

Telecom today has become an essence of life. Everywhere we see people using their smart phones for calling, checking email or accessing internet. To handle all these kinds of services without any intrusion is a very challenging task. This study deals with software testing which helps to ensure the quality of service to the end user. Software testing is an essential part in the software development process. Software development for telecom domain might not look as safety critical as of an airplane or nuclear reactor but it is arguably more complex. The main focus of this study is to provide automation to the unit testing of different types of radio connections that can be assigned to the end user based on the requested service and capacity of the 3G network. This research is sponsored by Ericsson to improve the testing of User Equipment Radio Connection Handling system that controls multiple possible radio connection configurations. This research attempts to identify and test all possible transitions between radio connection states. This will improve the existing manual state testing system, where changes in connection states cause dramatic impacts on test fixtures. As a solution, an automatic test case executor is proposed that generates possible transitions, which are later executed and verified automatically.

abstract: Healthcare operations have enjoyed reduced costs, improved patient safety, and innovation in healthcare policy over a huge variety of applications by tackling prob- lems via the creation and optimization of descriptive mathematical models to guide decision-making. Despite these accomplishments, models are stylized representations of real-world applications, reliant on accurate estimations from historical data to jus- tify their underlying assumptions. To protect against unreliable estimations which can adversely affect the decisions generated from applications dependent on fully- realized models, techniques that are robust against misspecications are utilized while still making use of incoming data for learning. Hence, new robust techniques are ap- plied that (1) allow for the decision-maker to express a spectrum of pessimism against model uncertainties while (2) still utilizing incoming data for learning. Two main ap- plications are investigated with respect to these goals, the first being a percentile optimization technique with respect to a multi-class queueing system for application in hospital Emergency Departments. The second studies the use of robust forecasting techniques in improving developing countries’ vaccine supply chains via (1) an inno- vative outside of cold chain policy and (2) a district-managed approach to inventory control. Both of these research application areas utilize data-driven approaches that feature learning and pessimism-controlled robustness. / Dissertation/Thesis / Doctoral Dissertation Industrial Engineering 2018

This study examined the impact of principals' data-driven decision-making practices on student achievement using the theoretical frame of Dervin's sense-making theory. This study is a quantitative cross-sectional research design where principals' perceptions about data were quantitatively captured at a single point in time. The participants for this study were 253 rural school principals currently serving in schools across Texas, and included both males and females across all ethnic groups, including white, African American, Hispanic, Asian, Native American and other. A developed survey instrument was administered to principals. The findings from the quantitative SEM analyses indicated that the Principal Uses Data to Improve Student Achievement latent variable (Factor 1) and the Principal and Staff Ability to Analyze Data to Improve Student Achievement latent variable (Factor 2) were significantly and positively associated with student achievement. Higher scores on these two latent variables were associated with better student achievement. There was no statistical association between the Principal Uses Data to Design Teacher Professional Development latent variable (Factor 3) and this target outcome. In total, the three latent variables accounted for 6% of the variance in student achievement (TAKS). When the campus level outcome was considered, no statistically significant associations between any of the latent variables and this outcome were evident. In total, the three latent variables accounted for less than 2% of the variance in campus level.

Data Week UPC 2021 - día 3 / Data WeeK UPC es un evento anual organizado por las Facultades de Negocios e Ingeniería, con el propósito de reunir a investigadores y expertos en la gestión empresarial para reflexionar acerca del papel de la Ciencia de Datos en la generación de valor en las organizaciones. Nueve expositores de distintas instituciones se unirán a las 4 fechas del Data Week 2021 este 23, 25, 26 y 27 de noviembre, para reflexionar acerca de los retos en el proceso de la transformación de datos para la toma de decisiones. No se pierdan la oportunidad de participar en este espacio en el que discutiremos las principales tendencias en cuanto a la aplicación de la ciencia de datos en la gestión empresarial. 7:00 PM EL CIENTÍFICO DE DATOS EN LAS ORGANIZACIONES DATA DRIVEN Ante la necesidad de manipular o manejar software, lenguajes especializados, capturar, procesar, analizar y representar grandes cantidades de datos, ¿cuáles son los perfiles requeridos para responder a esta nueva realidad? En esta charla abordaremos los perfiles relacionados con la ciencia de datos y su papel en el presente y el futuro de las organizaciones. 8:00 PM EL PAPEL DE LOS MODELOS PREDICTIVOS EN LA TOMA DE DECISIONES EMPRESARIALES Las organizaciones enfrentan muchos retos en el proceso de implementar modelos de machine learning y crear una cultura de data-driven. En esta charla se aborda el papel estratégico de los modelos predictivos para detectar riesgos en diferentes frentes. La aplicación de la ciencia en los datos permite un conocimiento cada vez mayor como base para la toma de decisiones empresariales."

This research focuses on the development of optimal sailing maneuvers for an AC75 foiling sailboat competing in the America's Cup. The America's Cup is the oldest, most prestigious, and technologically advanced sailboat racing competition in the world. Each iteration brings new and innovative sailboat designs which drastically improve sailing performance but increase complexity in the control of the sailboat system. This added complexity in the design and operation of the AC75 sailboat presents many challenges to the development of optimal sailing maneuvers. These challenges arise from the introduction of extra degrees of freedom and articulations in the boat such as the canting mechanisms (hydrofoils), which result in complex dynamical behaviors. The sailboat system is nonlinear, high-dimensional, and highly unstable. These complex characteristics require the development of high-order models, which are often intractable, or which introduce significant delays making them not well-suited for real-time control. The optimal maneuvers were achieved via the exploration of out-of-the-box solutions through data-driven controls and optimization. We used a high-fidelity sailboat simulator for the data generation process, and data-driven optimization schemes, such as Artificial Neural Networks (ANN), Extremum Seeking Control (ESC), and Jacobian Learning (JL) to optimize the sailing maneuvers. The optimizations were performed separately on various sailing maneuvers including close-hauled, tacking, and takeoff, as well as combinations of these maneuvers as performed during a race. The close-hauled and tacking maneuvers were optimized to achieve maximum Velocity Made Good (VMG) and minimum loss of VMG, respectively. The takeoff maneuver was optimized for maximum VMG and minimum time for the boat's transitions from displacement mode to foiling mode. The optimal solutions are subject to physical constraints and operational constraints enforced by the humans (sailors) in the loop. These maneuvers were developed for various heading angles (True Wind Angle (TWA)) and environmental conditions, such as True Wind Speed (TWS). Additionally, we performed an in-depth analysis of the optimal parameter settings obtained for close-hauled sailing to discern general trends in the parameter space. The trend of optimal parameters versus the wind direction provides a good understanding of the parameter space for varying sailing directions which can help guide the sailor's decisions during a race. The results show how optimization and controls can play a significant role in the development of advisory systems for complex human-operated systems. Lastly, the maneuvers developed in this search serve as performance benchmarks and provide insightful information about the underlying dynamics of the boat. / Mechanical Engineering