-
Prospective Students
-
Faculty and Staff
-
Current Students
-
Alumni and Friends
Family engagement is a crucial component of student success, impacting academic performance, attendance rates, and behavior. However, many families, particularly those from historically marginalized communities, remain disengaged from their child's school due to barriers such as a lack of trust, negative experiences, and language or cultural obstacles. A foundational reason for this disengagement is the unpreparedness of teachers to intentionally engage families. Teacher education programs often do not have an explicit focus on family engagement, resulting in teachers who may feel unprepared and who do not understand the cultural context of their students' families; thus, hindering effective communication. This dissertation explored the preparedness of beginning teachers to engage families in elementary schools, and how they perceive this preparedness, particularly in urban settings. By examining how beginning teachers perceive their readiness, it provided insights into the strengths and weaknesses of teacher education programs in this regard. The research sought to answer two central questions: 1) How are beginning teachers prepared to engage parents and families in elementary schools, and 2) How do they perceive their teacher education program's preparedness for this task? The study employed a mixed methods approach, involving curriculum analysis, online surveys , and semi-structured interviews. The findings of this study informed recommendations for teacher education programs, looking to equip future teachers with the skills and knowledge needed for effective family engagement.
Wastewater-based epidemiology (WBE) has emerged as a valuable tool for monitoring the spread of human respiratory viruses, particularly in the context of the COVID-19 pandemic. By bypassing traditional clinical testing, WBE can serve as an early indicator for viral outbreaks, enabling communities to make informed public health decisions. While WBE has been primarily used for SARS-CoV-2, its potential extends to other HRVs, including influenza A and B, and respiratory syncytial virus (RSV). In this study, we implemented a next-generation sequencing (NGS) protocol to assess human respiratory virus RNA in both wastewater and nasopharyngeal swabs that PCR tested negative for SARS-CoV-2. Control mixtures containing synthetic HRV RNA were spiked into wastewater and nuclease-free water to evaluate any matrix effects on sequencing outcomes. Bioinformatics analyses used taxonomic classification and direct alignment methods to compare the accuracy of human respiratory virus identification between wastewater and clinical samples. Despite the potential of NGS-based target-capture assays to detect viral genera, sequencing results from both wastewater and clinical samples demonstrated low depth and breadth of coverage, with discordant outputs from different bioinformatics pipelines. These findings highlight the need for rigorous benchmarking of laboratory and computational methods to ensure accurate human respiratory detection in wastewater and suggest that current sequencing approaches may fall short in providing the strain-specific information required for detailed public health surveillance.
The widely reported increase in the frequency of high impact, low probability extreme weather events pose significant challenges to electric power system's resilient operation. This dissertation research explores strategies to enhance operational resilience that addresses the distribution network's ability to adapt to the changing operating conditions. We introduce a novel Dual Agent-Based framework for optimizing the scheduling of distributed energy resources (DERs) within a networked microgrid (N-MG) using the deep reinforcement learning (DRL) paradigm. This framework aims to minimize operational and environmental costs during normal operations while enhancing critical load supply indices (CSI) under emergency conditions. Additionally, we introduce a multi-temporal dynamic reward shaping structure along with the incorporation of an error coefficient to enhance the learning process of the agents. To appropriately manage loads during emergencies, we propose a load flexibility classification system that categorizes loads based on its criticality index. The scalability of the proposed approach is demonstrated through running multiple case-studies on a modified IEEE 123-node benchmark distribution network. We also test the proposed method with different DRL algorithms to demonstrate its compatibility and ease of application. We compared the results with the traditional metaheuristic algorithms namely particle swarm optimization (PSO) and genetic algorithm (GA). To gain a deeper understanding of the developed model, we conducted a sensitivity study. The key findings from this study align with the mathematical foundation of the approach outlined in this dissertation, providing further support.
Diffusion is a scientific phenomena that can be modeled by partial differential equations. In this dissertation we first explore the development of equations for local, nonlocal, and quasi-nonlocal diffusion. Methods of finding solutions will be discussed as well as the properties of each diffusion model type. These properties include satisfying the maximum principle and demonstrating the well-posedness of each model which is through the solutions existence, uniqueness, and stability.
Also in a recent paper, a quasi-nonlocal coupling method was introduced to seamlessly bridge a nonlocal diffusion model with the classical local diffusion counterpart in a one-dimensional space. The proposed coupling framework removes interfacial inconsistency, preserves the balance of fluxes, and satisfies the maximum principle of the diffusion problem. However, the numerical scheme proposed in that paper does not maintain all of these properties on a discrete level. We resolve this issue by proposing a new finite difference scheme that ensures the balance of fluxes and the discrete maximum principle. We rigorously prove these results and provide the stability and convergence analyses accordingly. In addition, we provide the Courant-Friedrichs-Lewy (CFL) condition for the new scheme and test a series of benchmark examples which confirm the theoretical findings.
ABSTRACT
GRANT W. BIDNEY: Fabrication, Numerical Modeling, and Testing of Silicon Micropyramid Arrays and Retroreflectors
(Under the direction of DR. VASILY N. ASTRATOV)
This dissertation is devoted to the optical properties of mesoscale and nanoscale photonic arrays, specifically regarding two different areas: i) silicon (Si) micropyramidal photonics aimed at enhancing photodetectors and emitters, and ii) plasmonic Littrow retroreflectors in the optical regime.
In the first area, we show that Si anisotropic wet etching is attractive for the fabrication of large-scale arrays of micropyramids, or microvoids, with an extraordinary level of uniformity over centimeter-scale wafers. This is related to the self-terminating nature of the etching process when two (111)-type planes meet under the conditions when a surfactant is used to slow down the undercutting rate of the SiO2 layer. Although this technology is generally well studied by the microelectromechanical (MEMS) community, it seems that this particular property did not receive sufficient attention in previous studies. However, it is this property which enables the fabrication of uniform micropyramid arrays suitable for integration with detector and emitter arrays in optoelectronics applications. The optical properties of such arrays are studied by 3-D finite-difference time-domain (FDTD) numerical modeling in two realms represented by different boundary conditions (BCs). Periodic BCs result in Talbot self-images experimentally observed in this work. Perfectly matched layer BCs describe mesoscale interference effects resulting in the subwavelength focusing properties of individual micropyramids. It is proposed that integration with micropyramid arrays can enhance the collection of photons, signal-to-noise ratio, and operational temperatures of mid-wave infrared photodetector focal plane arrays (FPAs). It is also proposed that Si micropyramid arrays can be used to enhance light extraction and directionality of quantum sources and infrared scene projectors. Additionally, micropyramids were monolithically integrated with silicon-platinum silicide (PtSi/p-Si) Schottky barrier photodetectors to experimentally demonstrate an improved signal obtained by these micropyramid arrays. These results were compared with 3-D FDTD numerical modeling, as well as the modeling of a novel resonator cavity micropyramid structure as a way to further increase the enhancement capabilities of these micropyramids based on using a silicon-on-insulator (SOI) wafer. This structure demonstrated increased absorption of up to 11× compared to a planar reference device of the same size.
In the second area devoted to Littrow grating retroreflectors, we tackle the problem of simultaneous and efficient TE and TM polarization retroreflection. We developed the guidelines for designing such retroreflectors. Optimized performance at wavelengths in the vicinity of λ = 633 nm is expected for top metal slot arrays with thickness in the 20-40 nm range. However, this can vary for different metals such as Au, Ag, Al, and Cu. The most interesting development is our proposal to use the experimentally measured index values for thin films with different thicknesses to study and optimize the performance of real physical retroreflector devices. To the best of our knowledge this approach was proposed for the first time in our work. Using this approach, we showed that there is potentially plasmonic enhancement mechanisms involved, caused by their confinement in the metal stripes of the arrays. We demonstrated that, despite presence of absorption, such Au Littrow retroreflectors reach simultaneous ~0.2 and ~0.6 efficiency levels at TE and TM polarizations simultaneously in the same structure.
Heat shock protein 70 is an evolutionary conserved molecular chaperone responsible for the protein quality control functions. It is involved in many critical cellular processes, including folding protein ‘clients’, modulation of protein-protein interactions, and transport of proteins across membranes. Hsp70s are critical for maintaining cell viability in response to a large variety of cellular stresses. Perturbation of the proteostasis network is implicated in many diseases ranging from cancer and neurodegeneration to genetic disorders. Hsp70s are highly modified at the post-translational level. All these modifications together are referred to as the “chaperone code. These modifications fine-tune chaperone function, altering chaperone activity, localization, and selectivity. Understanding the regulation of these modifications will provide new insights into the protein folding process and characterize the direct interplay between chaperones and major signal transduction pathways. This thesis investigates a critical post-translational modification (PTM) site on yeast Heat Shock Protein 70 (Hsp70) that undergoes phosphorylation during heat shock response. Here, we focus on threonine 492 (T492), a highly conserved residue on Hsp70, which is conserved across all domains of life. Elucidating its upstream regulation and downstream effects. Yeast cells respond rapidly to heat stress by activating multiple protective mechanisms to maintain proteostasis. These include Hsf1 and Msn2/4-mediated transcriptional activation, cell integrity signaling, stress-induced bimolecular condensate formation and resolution, and protein translation inhibition. However, these pathways' rapid activation and coordination have remained poorly understood. Our findings reveal that heat-induced membrane stretch is detected by the Mechanosensor Mid2, triggering rapid phosphorylation of the cytosolic Yeast Hsp70 at T492. This phosphorylation event has several crucial downstream effects, which include altered interactome, altered dynamics of P-body resolution, maintenance of translational fidelity, amplification of the cell-wall integrity pathway, proper activation of heat-shock response, and regulation of clients Bck1 and Edc3. These results provide a comprehensive, unified theory of the global yeast shock response mediated by the Hsp70 chaperone code.
For centuries trusted advisors have helped leaders address knowledge gaps and provided an opportunity to evaluate logic processes and ideas before executing them. In industry, management consultants have turned the trusted advisor role into a profession that has increasingly garnered academic focus over time. While the benefits of management consulting may be difficult to quantify, the study of those benefits has been primarily case based and focused on publicly traded companies. Family businesses constitute 59% of the private sector workforce and 54% of private sector GDP in the US, representing a significant impact on the economy. But we know little about what influences a family business to seek external help or when a family business hires management consultants. The present study extended bounded systems theory to explore how family influence and succession intentions affect the intention to hire management consultants, and how performance aspirations moderate this relationship. The research identified a positive relationship between succession intentions and the intention to hire management consultants. It also demonstrated that family influence is not a statistically significant determinant of intention to seek external help. The results from this study help advance academic knowledge and provide useful insights to practitioners.
As cyber threats continue to grow in both volume and sophistication, automated and effective threat hunting has become essential for proactively detecting and responding to cyber threats. Unlike traditional defenses, an automated end-to-end threat-hunting approach involves analyzing vast amounts of unstructured data to identify actionable intelligence for timely detection and mitigation. Generative AI-driven threat hunting provides a more efficient and effective alternative due to the capability of understanding complex natural language patterns, enabling faster response times and greatly reducing human effort in identifying and analyzing threats. This dissertation aims to develop an automated end-to-end threat hunting model, harnessing the power of Large Language Models (LLMs) to enhance threat detection and response. This dissertation has three main objectives: 1) developing an approach to identify threat-related information from a large amount of unstructured text, 2) developing a model to extract actionable intelligence and explain them to gain the trust of security analysts, and 3) developing a model to generate search queries for log analysis, allowing security teams to investigate potential threats in a network.
The results of every step of the automated end-to-end threat hunting process have demonstrated the effectiveness of the approach. This dissertation achieved 94.93% precision and 88.22% recall in distinguishing between threat-related and non-threat-related real-time messages. The extraction step extracted critical threat information like IOCs, observable technical manifestations of attacks, and TTPs from threat-related messages. Additionally, by integrating a knowledge-graph based validation approach, the system ensured the accuracy of the extracted information and successfully reduced the hallucination rate from 34.6% to 1.58% and the error rate from 36.9% to 7.21%. Finally, this dissertation utilized the relational context in Kibana query generation and increased accuracy from 41.03% (without relational context) to 58.97%.
This dissertation presents several major contributions to automate the end-to-end threat-hunting process, transforming cyber threat intelligence messages into actionable Kibana queries to search logs for evidence of the attack described by the intelligence. This prototype implementation, leveraging OpenAI APIs, utilizes the robust language capabilities of Large Language Models (LLMs) to identify threat-related messages and extract actionable threat intelligence from even the most cryptic real-time threat-sharing messages. The core idea is to apply an explainable AI approach that explains the logical reasoning behind extracting threat intelligence, addressing a fundamental problem of LLM: hallucinations. This research explains the extracted intelligence in terms of specific MITRE ATT&CK TTP using a knowledge graph that includes “is-a” and “part-of” relationships, which are also extracted using a Large Language Model by OpenAI. The benefits of this explanation-based approach include significantly overcoming LLM hallucinations and gaining the trust of security analysts by providing explained results. Finally, this system leverages the explained “is a” and “part of” relationships to automate Kibana query generation for log analysis. This dissertation has demonstrated that explanation can improve LLM’s accuracy in generating Kibana queries and can be further extended by enriching the knowledge graph with additional relationships.
Medically unexplained symptoms (MUS), defined as symptoms lacking objective test findings or known biological causes, are highly prevalent and pose significant challenges for healthcare providers. Often associated with complex biopsychosocial origins, MUS can lead to diagnostic uncertainty. Consequently, providers may rely on patient characteristics, such as gender and mental health history, when making a diagnosis or determining appropriate treatments, which may introduce bias into their decision-making. This study investigated how these factors influence provider decision-making in diagnosing and treating MUS, focusing on two key research questions: (1) How does knowledge of a patient’s gender and mental health history affect diagnostic assessment? And (2) How does it impact treatment likelihood?
A sample of 152 primary care providers participated in the study, through an online survey, which implemented a 2x2 factorial between-subjects design. Participants were randomized into one of four conditions and reviewed clinical case vignettes, responding to questions regarding diagnostic and treatment considerations. The findings revealed a significant effect of patient gender and mental health history on treatment decisions. Providers were less likely to recommend medical follow-up for female patients with a history of depression and anxiety compared to male patients without a history of mental health concerns. For symptoms specifically involving generalized pain and fatigue, providers were more likely to attribute them to behavioral health factors than medical causes in female patients with histories of depression and anxiety compared to other groups. Conversely, for patients without a mental health history, providers favored medical follow-up over behavioral health interventions, regardless of patient gender. No significant differences emerged for diagnostic assessment or behavioral health treatment recommendations across groups.
These results suggest that patient gender and mental health history influence provider decision-making regarding the management of MUS, highlighting the need for strategies to reduce bias and improve equity in clinical decision-making. Additional research is warranted to explore these relationships further and better understand how various factors impact the assessment and treatment of ambiguous symptoms.
Transportation electrification is a critical global policy, and the effective integration of electric vehicles (EVs) is key to ensuring power grid reliability and resilience. This dissertation proposes solutions to address the integration of fleet EVs (FEVs) into distribution grids, using cloud-based and edge-based approaches.
First, a centralized, cloud-based optimization model is developed for the strategic placement of FEV charging stations, designed to enhance grid resilience during high-impact, low-probability events like hurricanes.
Second, a bilevel optimization framework is introduced to aggregate FEV charging with other distributed energy resources (DERs), enabling their participation in the ISO day-ahead energy and reserve markets. This approach models FEVs as part of a Distributed Energy Resource Aggregator (DERA), providing energy and ancillary services.
Third, a novel cloud-edge collaboration framework is proposed to enable decentralized control of FEV charging stations at the grid edge. This framework uses federated reinforcement learning, allowing individual FEVs to coordinate their actions locally while contributing to voltage regulation and grid stability.
These solutions offer a comprehensive approach for optimizing the deployment and control of FEV charging stations, addressing both operational efficiency and market integration in modern power grids.
