33760 research outputs found
Sort by
OVERT-COVERT MOVEMENT, COPY DELETION, AND CHAIN REALIZATION IN VALDÔTAIN PATOIS
No full textThis dissertation investigates patterns of chain realization in Valdôtain Patois, an understudied Francoprovençal language spoken in Aosta Valley (Italy), which permits some degree of optionality in wh-fronting: wh-phrases can occur sentence-initially or clause-internally (CIwh-phrases). Using several diagnostics (including binding and parasitic gap licensing), I argue that CIwh-phrases are in fact the result of movement in narrow syntax with deletion of higher copies in the chain. All wh-phrases move to their scope position in narrow syntax. The optionality arises post-syntactically, when the decision of which copy to pronounce is taken. Therefore, Valdôtain Patois constitute new evidence of the underexplored phenomenon of movement in narrow syntax with deletion of higher copies in the chain or overt-covert movement (Bobaljik, 1995, 2002;Bobaljik and Wurmbrand, 2012; Bianchi, 2019; Amaechi and Georgi, 2020). I argue that the patterns of optionality in copy pronunciation are reflexes of successive-cyclic movement, and follow from a specific and partly free ordering of (internal) Merge and upward Agree operations. Moreover, I discuss the pragmatic licensing of wh-questions in Valdôtain Patois. CIwh-
phrases are pragmatically marked: they act as strong presupposition triggers (Kripke, 2009; Abrusán, 2011; Abusch, 2010) and need contextual activation to be licensed. Therefore, we are faced with a very intriguing relation between the copy the grammar selects for pronunciation and the pragmatics of the question, which is problematic for more traditional models of grammar (e.g. the Y-model), where LF and PF are independent from each other and only get input from narrow syntax. In my analysis, I argue that these differences in use can still be accounted for by a purely syntactic feature-based account, with the pragmatic restrictions coming from an LF operator that restricts the possible answers to the question.
Finally, I discuss the island-insensitivity of CIwh-phrases and their matrix scope. Using interveners and parasitic gaps licensing, I show that wh-phrases move to the matrix CP area, but choosing to pronounce a lower copy in the chain voids the island effect. I adopt Fox and Pesetsky’s (2005a) account of successive-cyclicity as a linearization requirement and argue that, in Valdôtain Patois, choosing to pronounce a lower copy in the chain lifts the linearization requirements for subsequent steps of movement to happen successive-cyclically through each phase edge (including the island edge), and the island violation never occurs
Quantifying microplastics in water and sediment along river-marsh transects in the Choptank River
No full textPlastic began as useful material with many different applications. Due to its widespread production and consumption, its utility and durability are now also contributing to its status as a major environmental and human health contaminant with a long and mostly unknown lifecycle. This thesis first quantifies microplastics in the water and sediment of the Choptank River, a major tributary of the Chesapeake Bay. Microplastics are defined as particles smaller than 5mm that are produced as precursors to larger plastic material or occur when plastic items degrade. We explore the effect of differences in microplastic concentrations due to locations of transects along and positions across the river, seasonality, and interactions with vegetation. In the water column, abundance of microplastics was higher in the marshes flanking the river than the deeper channel at all transects and in all seasons. In the sediment, abundance is higher at subtidal than intertidal sites. The second part of the thesis explores the potentially novel microbial ecosystem(s) generated by the ubiquitous presence of plastic and the possibility of altering metabolic processes, subsequently restructuring biogeochemical flows
The Flugelhorn:
No full textAbstract
Title of Dissertation: The Flugelhorn: A Versatile Instrument for
Interchangeable Repertoire
Aunna V. Marzen, Doctor of Musical Arts, 2025
Dissertation directed by: Distinguished University Professor, Chris Gekker, School of Music
The concept of interchangeable repertoire is not a new idea to the music industry. Musicians from all backgrounds have chosen to perform repertoire from instruments that are not their own and often find success in doing so. The primary objective to this recording and dissertation project was to highlight the abilities of an instrument that deserves to be used in other diverse contexts beyond the standard jazz band and brass band settings. The flugelhorn not only processes a special ability to be diverse in what it can accomplish musically but also has a unique historic relationship to the euphonium and the saxhorn family. Taking advantage of the flugelhorn’s versatility, this project celebrates its ability to execute intricate euphonium repertoire and therefore adding additional repertoire the instrument’s classical perspective. Works include Sonata for Flugelhorn and Piano by Carson Cooman, “Psalm” from Two Portraits by Joseph Turrin, Sonata no. 1 for Trumpet and Piano, “Sonata for Heroes,” Movement II. With Integrity, by Marcus S. Grant, Fantasia, by Gordon Jacob, Suite for Horn and Piano, by Alec Wilder, and Concertino for Euphonium and Concert Band by Rolf Wilhelm
Accurate assessment and impairment-specific rehabilitation training with novel robotic devices for acute/sub-acute stroke survivors
No full textStroke is the main cause of adult disability in the US, with nearly 800,000 cases each year. With the rate of new cases increasing, there is a growing need for effective and precise rehabilitation. Robotic guided therapy has been employed ever more frequently in recent years. Although promising results have been achieved, there is more improvement needed to make robotic rehabilitation the new precision rehabilitation standard. We believe that significant gains could be achieved by conducting accurate assessment of impairment and initiating therapy during the acute phase post stroke, when the neuroplastic mechanism in the human brain is most effective. However, the absence of well‐defined protocols and clinic-ready devices has limited the delivery of early guided precision rehabilitation in the acute stroke care setting. In an effort to address these problems, in Chapter 2 we seek to accurately assess the complex changes that stroke induces across the multiple joints in the upper extremities, looking at passive, active and somatosensory properties. In Chapter 3 we will present the development of a new device, the FlexiArm an arm-hand exoskeleton designed for acute and subacute stroke rehabilitation. Finally, in the fourth chapter we will present a new facilitated neurorehabilitation strategy for acute patients post stroke based on our novel occlusion enhanced therapy. Through this work we have gained a deeper understanding the effects of stroke on the upper extremities. We have developed a series of rehabilitation robots designed to be portable and lightweight, suitable for acute stroke rehabilitation. We finally explored occlusion/reperfusion techniques applied to acute and subacute subjects and developed a new rehabilitation protocol to facilitate recovery of stroke survivors with severe impairment
Guiding Research, Building Belonging: Librarians' Role in Honors Undergraduate Research
No full textParticipation in high-impact educational practices (HIPs) such as living-learning programs (LLPs) and undergraduate research are continually cited as strong indicators of student engagement in higher education. The current literature shows students participating in HIPs exhibit a stronger sense of belonging on their campus. While academic libraries have regularly provided support to HIPs in higher education, much of the literature has focused on academic libraries as institutional support, instead of academic librarians as individual support. This case study of an undergraduate research-focused LLP at the University of Maryland, College Park investigated how mentorship from academic librarians is positively contributing to students' sense of belonging on campus.DOI: 10.4018/979-8-3373-0644-5.ch01
From Demonstration to Dynamic Interaction: Enabling Long-Term Robotic Planning
No full textRobotic learning has seen rapid growth over the past decade, driven by advances in machine learning that have brought real-world deployment of robots closer to reality. Research in this area primarily falls into two categories: reinforcement learning and imitation learning. Despite their promise, both approaches face significant challenges, including limited data availability and the difficulty of obtaining accurate state representations. This thesis explores how we can advance these methods to enable robust performance in real-world, unstructured environments.
We begin by exploring how to redefine state representation, presenting two complementary approaches. The first focuses on human state representation but is easily extendable to robots. It significantly outperforms existing methods in generalizing to unseen states and varying camera viewpoints. The second approach introduces a more concise, keypoint-based representation. We show that this method enables training of robot policies with minimal demonstrations and generalizes effectively to new environments and objects of varying shapes and sizes.
Next, we turn to the problem of learning policies from a single demonstration, without relying on handcrafted reward functions. Remarkably, our method achieves comparable final performance to existing approaches while using 100× less data. Finally, we demonstrate how these methods can be deployed in dynamic environments, even when trained under static conditions. By layering a lightweight planner on top of a pretrained policy, we achieve substantial improvements over naïve replanning strategies, approaching oracle-level success rates
SEQUENTIAL, HIERARCHICAL, AND ANALOGICAL PLAN TRANSFER IN ROBOTICS
No full textAs robotic systems encounter increasingly complex domains, efficient plan transfer—within and across distinct planning contexts—becomes essential to achieving adaptability and operational efficiency. This dissertation formalizes task-level plan transfer by introducing a conceptual and mathematical framework based on category theory, defining three types of transfer: sequential, hierarchical, and analogical plan transfer. For sequential and hierarchical transfers, symmetric monoidal categories (SMCs) and string diagrams are used to provide a rigorous framework for coherent task composition. To enable analogical transfers, a new planning representation language that integrates structured knowledge through -sets and double-pushout (DPO) rewriting is introduced. Functorial data migrations are then used to align and preserve semantic structure during analogical plan transfer, circumventing the need to re-plan. The efficacy of this framework is demonstrated through case studies across several applications, from industrial automation to service robotics, illustrating the broad applicability of these methods. In doing so, this work contributes to the growing body of research demonstrating how category theory unifies heterogeneous representations across computer science. Ultimately, this work advances AI planning and robotics by establishing a principled foundation for task-level plan transfer that enhances the safety, flexibility, and interpretability of robotic task planning in complex and knowledge-rich environments
Sources of Family Support and Their Association with Child and Family Well-Being
No full textAs both adverse and positive experiences in childhood can have enduring effects on health and well-being, identifying factors that promote optimal child development can have lifelong effects. Flourishing is a measure of overall well-being that includes positive mental health, emotional well-being, social behaviors and relationships. This study examines social supports that may enable families to thrive and to provide nurturing environments for child development, promoting flourishing even in challenging times. The study builds on prior research by examining social supports across socioecological levels and examining relationships between child and family well-being at the population level. This study found that children were more likely to flourish when their family had each of the social supports examined: emotional support, neighborhood support, and family-centered care. The prevalence of flourishing increased with more supports, and in all age groups children whose families reported all three of the supports had the highest increased likelihood of flourishing compared to children whose families reported no supports (aPRR = 1.29 for young children, 1.27 for school aged children, and 1.48 for adolescents). Additionally, there was a significant positive association between each of the social supports and family resilience, defined as the capacity of the family system to withstand and rebound from adversity. Children whose families reported all three of the supports had the highest increased likelihood of family resilience compared to those that reported no supports (aPRR = 1.41 for young children, 1.29 for school aged children, and 1.48 for adolescents). Further, family resilience partially mediated the associations between the supports and flourishing, with the proportion mediated ranging from 8.0-23.2%. Finally, the study found no association between living in states with more comprehensive family leave policies and both child flourishing and family resilience. The high degree of variability observed in the state-level prevalence of flourishing and resilience suggests the presence of other more salient predictors of well-being.
This study highlights the benefits of social supports for children and families, as well as the crucial role that families have in supporting child development and well-being across the lifespan. Policies and practices that support parents and promote family well-being and resilience are crucial for enabling all children to thrive. Further research can expand our understanding of individual variation in the use of supports and how the supports promote child and family well-being
RESOLVING STEEP GRADIENTS FOR PHYSICS-INFORMED NEURAL NETWORKS: RICHARDS’ EQUATION AND CONVECTION-DIFFUSION EQUATION
No full textSolving partial differential equations (PDEs) is fundamental to modeling physical, biological, and engineering systems. Traditional numerical methods, while accurate and robust, often struggle with high-dimensional problems, complex geometries, discontinuous boundary conditions, and steep solution gradients. Deep learning, and in particular Physics-Informed Neural Networks (PINNs), offers a promising mesh-free alternative by embedding physical laws directly into the training process of neural networks.
This dissertation investigates and improves the performance of PINNs for PDEs with sharp transitions and nonlinearities. We focus on two key equations: Richards’ Equation, which models unsaturated flow in porous media, and the Convection-Diffusion Equation, which governs advective-diffusive transport processes. For Richards' Equation, we introduce a PINN framework that includes surface flux as an input and discrete residuals to enforce causality. For convection-diffusion problems, we address both known and unknown gradient layer locations. When the location is known, we analyze PINN limitations and propose input transformations that focus model capacity on high-gradient regions. When the location is unknown or complex, we introduce the Auxiliary-Input PINN (AI-PINN), a novel architecture that adapts spatial transformations based on an auxiliary input. A successive training strategy is also used to learn solutions without prior knowledge of gradient positions.
The methods developed in this work offer general strategies for improving the accuracy and robustness of PINNs when applied to challenging PDE problems with steep gradients. Through a combination of architecture design, loss reformulation, and training strategies, this dissertation contributes toward making physics-informed learning a viable tool for complex real-world simulations
MODELING CONTROL STRATEGIES OF A HIGH-PERFORMANCE ENERGY RECOVERY VENTILATOR IN NIST'S NET ZERO ENERGY RESIDENTIAL TEST FACILITY
No full textWith the acceleration of climate change causing increased global temperatures, reduced polar ice caps, and more severe weather, reducing carbon emissions is more important than ever. Residential buildings are responsible for 57 % of the greenhouse gas emissions in the building sector. This sector provides a clear opportunity to reduce emissions and energy consumption. Net Zero Energy Buildings (NZEB) are a more recent answer to this need where the building generates more energy over a year than it consumes through renewable sources. NIST’s Net Zero Energy Residential Test Facility (NZERTF) was constructed and instrumented to serve as a test bed for NZEB research at the residential scale. This facility has a previously uncharacterized Energy Recovery Ventilator with opportunities to implement a bypass mode to avoid any heat exchange between the fresh and stale air streams. This bypass mode allows for free cooling in the shoulder seasons by using cooler outdoor air to cool the inside of the building. To test the energy savings potential for temperature-based and enthalpic-free cooling control schemes, a TRNSYS model was updated and tuned to accommodate more recent building performance and shifting operation patterns. These updates included heat pump performance curves, adjustments to HVAC airflow modeling, moisture capacitance models, and a complete evaluation of the ERV’s fan power and sensible and latent effectiveness. This model was run for annual simulations without bypass mode, with bypass mode enabled based on the outdoor temperature, and with bypass mode based on the outdoor enthalpy. The temperature-based control required an additional 2.4 % of heat pump energy relative to the baseline along with a slight degradation in thermal comfort. The enthalpic control saved 0.2 % and maintained a similar thermal comfort to the no bypass mode case. These bypass controls do not offer significant opportunities for this facility in this climate zone but could be implemented in lower-humidity climates or more extensive facilities with more substantial sensible loads. I plan to continue this work by testing the enthalpy bypass control at the NZERTF, characterizing, modeling, and testing bypass options with the ventilator’s HRV core, and by integrating CONTAM into the TRNSYS model to improve the airflow modeling for more accurate zone level conditions