62588 research outputs found
Sort by
Modeling Excited States and Charge Transport with Density Functional Theory
No full textKohn-Sham density functional theory (DFT) and its periodic extension are widely used to investigate the electronic structure, energetics, and geometries of molecules and materials. While DFT reliably predicts ground-state properties, it struggles to accurately describe vibrational and electronic excited states as well as charge transport. Perturbative extensions, such as coupled-perturbed Kohn-Sham (CPKS) for infrared and Raman spectra and linear-response time-dependent DFT (LR-TDDFT) for UV-Vis spectra, attempt to address these limitations but remain constrained by system size and are typically restricted to vertical excitations. This thesis first introduces ground- and excited-state electronic structure methods (Chapter 1) and evaluates the accuracy of excited-state methods (Chapter 2). Ground-state isomer difference spectra were used to characterize small populations of pyridine-based azo dyes with DFT and LR-TDDFT (Chapter 3). A new approach is developed for predicting excited-state absorption spectra and interpreting transient absorption spectra by applying the linear-response Tamm-Dancoff approximation to non-Aufbau configurations (Chapter 4). To study charge transport, a finite-displacement method is introduced to quantify the impact of vibrational modes on carrier mobility in organic crystals (Chapter 5). Anisotropic carrier mobilities in organic crystals are subsequently predicted using Boltzmann transport theory (BTE, Chapter 6), and this framework is extended by incorporating the finite-displacement method to evaluate the detrimental effects of vibrational modes on charge transport in tetracene (Chapter 7). The methods developed in this work provide new insights into the photophysics of molecular chromophores and the role of vibrational disorder in charge transport and will continue to advance the theoretical understanding of these processes.</p
The Asteroseismology of Evolved Solar-like Oscillators in Eclipsing Binaries
No full text{"value":"Asteroseismology is invaluable for revealing the properties of stellar interiors and as a tool for determining bulk physical stellar properties. The characteristics of solar-like oscillations, specifically the maximum frequency and large frequency separation of pressure mode pulsations are directly related to the parameters of the star. This relationship led to the derivation and continued use of the asteroseismic scaling relations, which provide a simple way to determine stellar mass, radius, and age, all of which are essential parameters in testing current stellar evolution theory. However, the scaling relations are calibrated with solar values, and applying them to more evolved stars is known to be less accurate. Therefore, there is a need to re-calibrate the scaling relations for evolved solar-like oscillators. To do so, we may look to eclipsing binary systems containing an evolved component which exhibits solar-like oscillations. Stellar masses and radii in these systems can be precisely measured by modeling the eclipsing binary, and subsequently compared with the mass and radius found by applying the scaling relations. We have developed a pipeline (which will be made publicly available for community use) which can be used to search for solar-like oscillations in eclipsing binaries; the pipeline detrends observations and removes the eclipsing signal, as well as instrumental variability and star spot variations, from the light curve in order to detect power excess due to pulsations. We applied this pipeline to light curves of eclipsing binaries observed by the Transiting Exoplanet Survey Satellite (TESS) in search of targets with an evolved pulsating component, and produced a catalog of the targets found.
We have begun follow-up observations and analysis of several of these targets as a first step towards testing the scaling relations. Our catalog provides a valuable sample of systems which will be used to re-calibrate the asteroseismic scaling relations for evolved stars.
We also model an eclipsing binary with a red giant component which exhibits solar-like oscillations using the PHysics Of Eclipsing BinariEs (PHOEBE) software to measure the parameters of the stellar components and compare with the results using asteroseismology. This provides a first case study and framework for analyzing the rest of the targets identified in this work. ","attr0":"abstract"
Green Roofs as a Policy Solution for Large Warehouses
No full textLarge warehouses pose a challenge to the small municipalities in which they wish to locate. Externalities create problems for the municipalities in which warehouses are located. This study examines four of these buildings to identify mitigation methods, monitoring tools, and policy levers for managing the externalities created by large warehouses. Green roofs are a solution that counteracts excess heat. I suggest them as an appropriate way to mitigate heat in climates similar to Pennsylvania’s. They have been successfully used in the past to combat increased stormwater runoff. Policy tools such as incentives, regulations, fees, and fines that cities use to manage large buildings can be used by municipalities to help manage warehouses.</p
Bedtime Resistance and Sleep Habits in Preschoolers At-Risk for ADHD: Effects of Behavioral Parent Training
No full textSleep is a significant challenge for preschoolers with attention-deficit/hyperactivity disorder (ADHD) as they may experience bedtime resistance, difficulty initiating and maintaining sleep, daytime sleepiness, and medical sleep disorders. Additionally, sleep problems can exacerbate and be exacerbated by ADHD symptomology, leading to functional impairment. Behavioral parent interventions are an effective treatment for reducing ADHD symptoms among preschoolers and are similarly recommended for sleep problems. The present study examined the effect of behavioral parent training (BPT) for preschoolers with ADHD on bedtime resistance and sleep habits, comparing face-to-face and online intervention administrations, as well as overall BPT and wait-list control conditions. First, it was hypothesized that changes in bedtime resistance and sleep habits from pre- to post-treatment would not significantly differ based on BPT administration format. It was also hypothesized that receipt of BPT would improve bedtime resistance and sleep habits from pre- to post-treatment to a greater degree than for wait-list controls. Furthermore, the relation of pre-treatment bedtime resistance to post-treatment sleep habits and the moderating influence of BPT was explored. Data were collected as part of Promoting Engagement with ADHD Pre-Kindergartners (PEAK), a randomized controlled trial of a BPT program. Participants included caregivers of children ages 3- to 5-years-old with ADHD (analytic sample n = 109 families; children: Mage = 53.71 months, SD = 9.17, 69.7% male; caregivers: 86.2% female). Multivariate analyses of variance and moderated mediation analyses were conducted. When comparing face-to-face and online BPT groups and BPT and wait-list control groups, no significant main effects of group assignment, time period, or their interaction on bedtime resistance and sleep habits were found. Complete mediation was found for the relationship between pre- and mid-treatment bedtime resistance and post-treatment total sleep habits and sleep initiation habits. However, group assignment was not a significant moderator, thus moderated mediation was not present. Notably, several nonsignificant results were in the hypothesized direction, possibly limited by reduced power. Overall, findings may indicate BPT\u27s potential for supporting bedtime behavior and other behavioral sleep challenges among this population. Future studies should utilize objective sleep measures with larger samples and examine effects of supplementing BPT with sleep instruction. </p
An Exploration of Minority Stress for Sexual and Gender Minority Survivors of Intimate Partner Violence: The Role of Informal Systems of Support
No full textLesbian, Gay, Bisexual, Trans*, and Queer (LGBTQ+) people are at higher risk for psychopathology such as depression, anxiety, and posttraumatic stress disorder (PTSD). The minority stress theory (MST) explains these disparities as outcomes of external and internal identity-based oppression. Intimate partner violence (IPV) is also a greater risk for individuals in the LGBTQ+ community, and this experience, combined with minority stress, may contribute to unique and cumulative psychological distress. Informal systems of support, such as social support and connection to the LGBTQ+ community, might explain the relationship between minority stress and psychopathology. However, few studies examine sexual and gender minorities (SGM) in the same study, and even fewer focus specifically on the role of support for SGM IPV survivors. It was anticipated that this combination of foci on SGM people and their social networks would provide information that is protective for SGM survivors while highlighting the unique strengths of supportive SGM communities. A sample of 274 SGM IPV survivors was collected, and results were supportive of a minority stress model with distal and proximal stress as predictors of informal systems of support and psychopathology. LGBTQ+ discrimination, internalized transphobia, and family support were the strongest predictors of psychopathology, and family support mediated relationships between minority stressors and psychopathology. Findings from this study highlight the importance of reducing the existence, and impact, of identity-based oppression for SGM IPV survivors through research, clinical work, and advocacy.</p
Entrepreneurial Mindset for All: An Exploratory Comparative Case Study of Two-Year and Four-Year Entrepreneurial Education Courses
No full text{"value":"Abstract The concept of an entrepreneurial mindset is not new. Dweck (2006) initially popularized the concept detailing the difference between fixed and growth mindsets. A growth mindset lends itself to creativity, flexibility, and adaptability–in the workplace, in academe, and in all areas of life. But is a growth, or an entrepreneurial mindset, something that can be taught and learned?
The concept of “mindset” exists among entrepreneurship educators within higher education institutions worldwide (Casulli et al., 2022; Mawson et al., 2022). However, according to Larsen (2022), fostering entrepreneurship education in colleges and universities presents challenges developing instructional strategies promoting an entrepreneurial mindset amongst students. Fretschner and Lampe (2019) assert that one of the purposes of entrepreneurship education is the interest in an entrepreneurial mindset as a learning outcome. According to Nabi et al. (2007), the need exists for a better understanding of pedagogies in entrepreneurship education courses. Collins et al. (2004) suggest that gaps exist between student needs and higher education capabilities related to entrepreneurship education. Studying how instructional strategies of entrepreneurship education impacts various aspects of an entrepreneurial mindset is necessary (Larsen, 2022; Mawson et al., 2022). Thus, a critical need exists for educators to understand how instructional materials and delivery of entrepreneurial education courses can truly instill entrepreneurial mindsets to students in their classrooms.
This exploratory comparative case qualitative case study examined how instructors at a public two-year community college and a private four-year university employed teaching methods, instructional materials, and styles to deliver introductory entrepreneurial education courses. Through examination of course materials, class observations, and instructor interviews, it compared strategies utilized by the instructors to determine how entry-level entrepreneurial education courses were taught. The study examined instructor styles, materials, and teaching strategies to discover differences between two-year and four-year higher education instructors attempting to teach and foster entrepreneurial mindsets amongst their students. Findings suggest that a blended approach combining non-traditional reading materials that convey theoretical practices along with project-based group work could optimize the teaching of entrepreneurial education courses. Furthermore, it is necessary to examine non-traditional pedagogical teaching strategies modeling the philosophy of an entrepreneurial mindset itself.
","attr0":"abstract"
Finding Yourself through Empowerment, Raza, and Agency (FUERZA): A Critical Consciousness Mental Health Intervention for Latinx Youth
No full textLatinx immigrant youth face systemic stressors in many areas, including the sociopolitical climate, school climate, and family- and individual-level discrimination. These stressors are associated with poor academic achievement and low mental well-being among Latinx youth. Critical consciousness has long been used as a way to increase advocacy among marginalized youth and help them persist through systemic inequities. Additionally, liberation psychology as a framework, which includes critical consciousness, is culturally responsive to Latinx youth, as it provides a focus on their cultural strengths and their inherent ancestral wisdom to help them move through mental health hurdles. However, no studies to date have created a liberation psychology-informed critical consciousness intervention for Latinx middle school youth, specifically to improve their mental health. The current study created, piloted, and tested the effects of a five-session culturally-responsive preventative mental health intervention for Latinx middle school students, called FUERZA (strength in Spanish). This study used a pre- post experimental design, where 44 Latinx seventh grade students were non-randomly assigned by class in FUERZA (n = 23) or in the control group (n = 21). Results demonstrated that students in the treatment group had greater psychological well-being compared to the control group at post-test. Thus, FUERZA may have provided a buffer for psychological well-being decline that both groups experienced. Findings provide initial evidence on the efficacy of FUERZA. Implications and future directions are discussed for school leaders and school practitioners.</p
A Reinforcement Learning Approach to Burn Control in ITER
No full textThe highly nonlinear and coupled dynamics of burning plasmas in ITER will demand burn control, which is the active regulation of plasma temperature and density. Recently, reinforcement learning (RL) has been explored as an alternative to traditional model-based control synthesis for tackling plasma-control problems in tokamaks. Though guaranteeing steady tracking and convergence can be challenging, RL-based control synthesis may offer the capability to handle greater complexity in the plasma-response model used during training. In this work, an RL-based burn controller is designed to track user-specified, time-varying references for the plasma states. The reference-tracking controller is trained within a simulated burning plasma environment using a model-free reinforcement learning algorithm. This environment is modeled using a nonlinear, zero-dimensional plasma simulator called COBALT, Control-Oriented Burning plAsma simuLaTor, which captures the energy and density evolution in ITER. The burn controller adjusts six action inputs: external deuterium pellet injection, deuterium-tritium pellet injection, ion cyclotron resonance-heating, electron cyclotron resonance-heating, and two neutral beam injectors. It aims to minimize the discrepancy between the observed plasma state and the user-specified reference. The effectiveness of the proposed controller is demonstrated using closed-loop simulations based on ITER scenarios.</p
What Are Topic Modeling Metrics Measuring? A Comparison of Topic Modeling Metrics and Human Assessment Approaches
No full text{"value":"Evaluation of topics, topic labels and topic modeling techniques is often done via computational metrics or human-centered single item assessment. While prior work has shown some of the computational metrics correlate with human assessment, it remains underexplored whether human perception of the topic quality aligns with metric-based topic evaluation.This dissertation explores the evaluation of topics and topic labels, comparing computational metrics with human-centered assessment approaches.
It investigates whether human assessors consider multiple dimensions of quality beyond label goodness and explores the relationship between human assessment dimensions and common topic quality metrics.
Furthermore, the dissertation examines the alignment between how human users compare topic models and both aggregation of computational metrics and human assessment of individual topics. It investigates whether models favored by metrics or human evaluation of individual topics are also preferred by human participants.
The findings of this dissertation show that only a limited correlation exists between human assessment of individual topics and topic labels and NPMI metrics among other computational metrics. In addition to that, this dissertation reveals that aggregation of topic assessments with respect to metric evaluation or human assessment does not align with human comparison of different sets of topics.
This dissertation concludes by comparing human perception of topic quality across human participants with varying familiarity with the presented context and the task. It shows that higher familiarity can influence the perception of quality and human participants with a higher familiarity to the shown context or taken task assess quality of topics or topic labels in a way that the differences are more visible. ","attr0":"abstract"
A Learning-Based Decentralized Approach to Eigenspectrum Optimization of Network Graphs
No full text{"value":"This thesis presents a novel learning-based decentralized approach to eigenspectrum optimization for network graphs, aimed at mitigating vulnerability to adversarial resonance attacks. Traditional centralized methods, while effective, facescalability and feasibility challenges in large or decentralized networks. To address these limitations, we propose a deep neural network (DNN) model trained on localized outcomes from centralized eigenspectrum optimization. The model learns to predict edge weight adjustments based on local graph topologies, enabling decentralized control without global network knowledge. Our methodology involves generating diverse graph datasets, embedding maximal subgraphs to standardize
inputs, and iteratively refining the DNN architecture for optimal performance. Experimental results demonstrate that the decentralized approach achieves significant spectral flattening, reducing network vulnerability while adhering to constraints
such as fixed total edge weight and preserved topology. The Decentralized Optimization Performance Ratio (DOPR), a novel evaluation metric introduced in this thesis, quantifies the effectiveness of our method, revealing strong performance,
particularly in densely connected graphs. This work bridges the gap between centralized and decentralized optimization, offering a scalable and practical solution for enhancing network resilience against adversarial attacks.","attr0":"abstract"