The present study provides an example of implementing the difference in differences (DD) estimator for a two-group, pretest-posttest design with K-12 educational intervention data. The goal is to explore the basis for causal inference via Rubin's potential outcomes framework. The DD method is introduced to educational researchers, as it is seldom implemented in educational research. DD analytic methods' mathematical formulae and assumptions are explored to understand the opportunity and the challenges of using the DD estimator for causal inference in educational research. For this example, the teacher intervention effect is estimated with multi-cohort student outcome data. First, the DD method is used to detect the average treatment effect (ATE) with linear regression as a baseline model. Second, the analysis is repeated using linear regression with cluster robust standard errors. Finally, a linear mixed effects analysis is provided with a random intercept model. Resulting standard errors, parameter estimates, and inferential statistics are compared among these three analyses to explore the best holistic analytic method for this context.

<p dir="ltr">The rising demand for EVs, motivated by their environmental benefits, is generating increased need for EV charging infrastructure. Also, it has been recognized that the adequacy of such infrastructure helps promote EV use. Therefore, to facilitate EV adoption, governments seek guidance on continued investments in EV charging infrastructure development. The high cost of these investments motivates governments to seek optimal decisions on EV-related investments including EV charging infrastructure, and such decisions include locations, capacities, and deployment scheduling of such infrastructure. Additionally, uncertainties in travel demand prediction and EV driving range constraints need to be considered in EV infrastructure investment planning. To help address these questions, this thesis developed a framework to establish optimal schedules and locations for new charging stations and for decommissioning gasoline refueling stations for any given network over a long-term planning horizon, considering uncertainties in travel demand forecasts and EV driving-range heterogeneity. To address the uncertainties, the proposed framework is formulated as a robust mathematical model that minimizes the worst-case total system travel cost and the total penalty for unused charging station capacity. This study uses an adaption of the cutting-plane method to solve the proposed model. In the numerical analyses, the performance of the robust framework and its deterministic counterpart are compared. The results show that the optimal robust plan outperforms the deterministic plan by yielding savings in the costs of travel and electricity charging. The thesis also investigates the effects of investment budget levels of robust planning. The numerical results throw light on the relationships between higher investment levels and electric charging station deployment levels and consequently, the savings in travel costs and impacts on unused charging capacity. The outcomes of this thesis can help road agencies and related private sector entities enhance preparations towards infrastructure investments to support electric charging stations in an efficient manner.</p>

Umeå University Hospital has acquired a new treatment planning system, RayStation, for radiotherapy. It has a different set of dose calculation algorithms that require new planning strategies for stereotactic lung cancer treatments. This \paper{} has two parts and the first investigates the Monte Carlo algorithm, recalculating doses for 21 historical treatments, and finds that it is slow but fast enough to be used and that it reports dose coverage that is usually lower than that of simpler algorithms, particularly for small tumours in low density lungs. The second part investigates the performance of four alternative planning methods: a straight forward PTV/ITV based method, a method using density override of the PTV, an ITV based semi-robust method and a fully robust method treating both setup margins and breathing motion robustly. The new methods are tested on 12 historical cases and are compared, using robust evaluation, to each other and the original plans. The first three are found to all be viable alternatives, while the fully robust method, while impressive, is found not to be a practical alternative at this time. / Norrlands Universitetssjukhus har skaffat ett nytt dosplaneringssystem, RayStation, till radioterapin. Det har en annan uppsättning dosberäkningsalgoritmer som kräver en ny planeringsstrategi för stereotaktisk lungcancerbehandling. Den här rapporten innehåller två delar och den första undersöker Monte Carlo-algoritmen genom att räkna om doserna för 21 tidigare behandlingar. Den visar sig vara långsam men snabb nog för att vara användbar och den rapporterar lägre dostäckning än enklare algoritmer, särskilt för små tumörer i lungor med låg densitet. Den andra delen undersöker fyra alternativa planeringsmetoder: en baserad på PTV och ITV utan krusiduller, en dito med s.k. override av PTVs densitet, en ITV-baserad med robust hantering av setup-marginaler och en helt robust metod med robust hantering av både marginaler och andningsrörelser. De nya metoderna testas på 12 tidigare behandlingar och jämförs, medelst robust utvärdering, med varandra och originalplanerna. De tre första visar sig alla vara goda alternativ medan den helt robusta, som visserligen är imponerande, inte är praktiskt användbar än.

Soft robots have been used in a wide range of applications from robotic and mechanical engineering to medicine and biomededical field. The growing interest in soft robots comes from their good performance in environments which is not best suited for conventional rigid bodies. Soft robots utilize the compliance, adaptability and flexibility of soft materials and actuation methods to develop highly adaptive structures. Among the soft materials, elastomers are specially popular due to their wide range of elasticity and viscoelasticity. Along with elastomers, textile fabrics are also of high interest for soft robotic applications due to their bendable, flexible, and often stretchable nature. The reinforcement of elastomers with textile fibers results in so-called integrated fiber-elastomer composites (IFEC) which offer a wide variety of properties such as flexibility, strength, fracture toughness and damage resistance. The elastic properties of textile reinforced composites require smart actuators which possess adaptability and deformability. Among existing smart actuators, shape memory alloys (SMA) have been frequently adopted in flexible structures including soft robots. SMAs have sensing and actuation capabilities and are characterized by flexibility and lightness which facilitates their integration into these structures. In this dissertation, the modeling and control of soft prototypes made of IFEC are presented. Shape memory alloys are embedded in the composites for the system actuation. First, the mechanical design and production of three IFEC prototypes are described. For each prototype, a test bench including power and control electronics set-up is designed. Next, mathematical models are developed to analyze the dynamic behavior of the prototypes. The IFEC systems exhibit highly nonlinear behaviour due to SMA hysteresis. For modeling, two different approaches, namely physical modelling and system identification are adopted. In physical modeling, the SMA constitutive and heat transfer equations are incorporated with the composite deflection model. To fully develop the equations, thermal and mechanical parameters of SMA wires are identified experimentally. In the second approach, the mathematical model of the systems is derived from experimental identification and unstructured uncertainty models. Two different control techniques are proposed to compensate the nonlinear behavior of the systems and ensure a robust, fast and precise position tracking. In the first control technique, a proportional integral (PI) controller is designed through robust stability analysis. The second controller is a multivariable PI control which is designed for the prototypes that can move in more than one direction. The performance of the controllers are examined experimentally.

京都大学 / 新制・課程博士 / 博士(工学) / 甲第25013号 / 工博第5190号 / 新制||工||1991(附属図書館) / 京都大学大学院工学研究科機械理工学専攻 / (主査)教授 泉井 一浩, 教授 小森 雅晴, 教授 井上 康博 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

Over the past few years, speech recognition technology performance on tasks ranging from isolated digit recognition to conversational speech has dramatically improved. Performance on limited recognition tasks in noiseree environments is comparable to that achieved by human transcribers. This advancement in automatic speech recognition technology along with an increase in the compute power of mobile devices, standardization of communication protocols, and the explosion in the popularity of the mobile devices, has created an interest in flexible voice interfaces for mobile devices. However, speech recognition performance degrades dramatically in mobile environments which are inherently noisy. In the recent past, a great amount of effort has been spent on the development of front ends based on advanced noise robust approaches. The primary objective of this thesis was to analyze the performance of two advanced front ends, referred to as the QIO and MFA front ends, on a speech recognition task based on the Wall Street Journal database. Though the advanced front ends are shown to achieve a significant improvement over an industry-standard baseline front end, this improvement is not operationally significant. Further, we show that the results of this evaluation were not significantly impacted by suboptimal recognition system parameter settings. Without any front end-specific tuning, the MFA front end outperforms the QIO front end by 9.6% relative. With tuning, the relative performance gap increases to 15.8%. Finally, we also show that mismatched microphone and additive noise evaluation conditions resulted in a significant degradation in performance for both front ends.

No description available.

No description available.

No description available.

No description available.