Digital Commons Network^™

Large Language Models (Llms) For Clinical Note Generation: International Classification Of Disease (Icd) Code, Knowledge Graph (Kg) And Prompt Evaluation, Ivan P. Makohon

Computer Science Theses & Dissertations

In the past decade, a surge in the amount of electronic health record (EHR) data in the United States occurred, driven by a favorable policy environment created by the Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009 and the 21st Century Cures Act of 2016. Clinical notes for patients’ assessments, diagnoses, and treatments are captured in these EHRs in free-form text by physicians, who spend a considerable amount of time entering them. Manually writing these notes is time-consuming, increasing patient waiting times and potentially delaying diagnoses. Large language models (LLMs), such as GPT-4o, possess the ability …

On The Scalability Of Anisotropic Mesh Adaptation On Distributed And Shared Memory Architectures For Numerical Approximations, Kevin Mark Garner Jr.

Computer Science Theses & Dissertations

Mesh generation is a critical component in numerical approximations of Partial Differential Equations (PDEs). One such example includes Computational Fluid Dynamics (CFD), as CFD simulations in turn are crucial for applications in many industries, such as personalized healthcare and the design of aerospace vehicles. Generating high quality meshes for large-scale CFD problems presents a significant bottleneck in the CFD workflow. This dissertation proposes “fast,” parallel 3D mesh generation methodologies that are designed to leverage the concurrency offered by emerging High-Performance Computing (HPC) architectures. First, a distributed memory method is presented that integrates a sequential state-of-the-art isotropic, advancing front local reconnection-based …

The Paradox Of Discipline: Equity, Standardization, And Black Student Outcomes In Public Schools, Latricia T. Davenport

Educational Leadership & Workforce Development Theses & Dissertations

Despite efforts to address profound discipline disparities impacting Black students, the implementation of alternative disciplinary approaches like Positive Behavioral Interventions and Supports (PBIS) and Restorative Justice (RJ) has failed to eliminate significant disproportionality in school discipline. This study investigates how teachers navigate the challenges of disciplining Black students within a paradoxical school environment that juxtaposes equitable, human-centered methods with a historically rooted managerial paradigm emphasizing uniformity and zero tolerance. Drawing upon systemic racism and the concept of 'apprenticeship of observation,' this study explores how teachers' implicit biases about Black students are formed by long-held institutional norms and personal experiences. It …

Evaluation Of Ocean Lidar Enhancement Through Miniaturization And Gated Pmt-Based Range Extension, Chandler Austin Slater

OES Theses and Dissertations

Oceanographic lidar systems remotely characterize the vertical structure of the upper ocean by recording the light backscattered from a laser pulse as it propagates through the water column. Historically these systems have been constrained by limited detection ranges, often capturing only a portion of the illuminated water column, typically fading to noise around the 10% isolume. A primary constraint has been the limited dynamic range of digitizers, which inhibits the ability to resolve both intense near-field and faint far-field backscatter signals within a single acquisition. To address this, we developed a novel optical lidar system incorporating gated photomultiplier tubes (PMTs) …

Aluminum As A Tracer Of Dust Deposition To The Ocean: A Case Study From The Bermuda Region, Tara Elizabeth Williams

OES Theses and Dissertations

Aluminum (Al), a major component of mineral aerosol (dust), partially dissolves in seawater and is widely used as a tracer for estimating time‐averaged dust fluxes to the ocean. Such estimates rely on dissolved Al (DAl) inventories in the surface mixed layer (SML), an assumed SML residence time of DAl (_TDAl), the fractional solubility of Al in dust (Al_S), and the mass fraction of Al in dust. In this study, dust flux estimated from seasonal, water-column DAl data from the Bermuda Atlantic Time-series Study (BATS) region are compared with direct dust flux estimated from contemporaneous measurements of …

Machine Learning For Anomaly Detection In Neural Network Security And Srf Cavities, Hal Ferguson

Electrical & Computer Engineering Theses & Dissertations

This dissertation explores the development and deployment of machine learning approaches to address critical challenges in anomaly detection across two distinct domains: neural network security in federated learning settings and cavity behavior analysis in particle accelerator operations at Jefferson Lab in Newport News, Virginia. Anomaly detection identifies deviations from expected patterns, safeguarding systems in cybersecurity, industry, and research against malicious activities and failures. This dissertation demonstrates how our machine learning approaches enhance detection accuracy and efficiency in both neural network security and industrial applications.

First, we investigate vulnerabilities in deep neural networks deployed in federated learning. Although federated learning preserves …

Relativistic Three Particle Scattering, Md Habib E Islam

Physics Theses & Dissertations

In this dissertation, we develop and apply a relativistic framework for studying three-body scattering amplitudes, their analytic structure, and the emergence of universal phenomena such as the Efimov effect. The work integrates three complementary components.

First, we introduce a systematically improvable numerical method for solving relativistic three-body integral equations in momentum space. By discretizing the continuum problem into finite matrix equations and extrapolating to the continuum limit, we obtain stable partial-wave amplitudes in the presence of a two-body bound state. Two complementary treatments of the pole contribution are implemented and shown to reproduce previous finite volume results, with controlled estimates …

A Kinematic, Data-Parameterized Approach To Constrain Subducted Seafloor During The Emplacement Of Shatsky Rise, Alexzandrya Katrina Shotorban

OES Theses and Dissertations

This study investigates ridge spreading rates and plate boundary geometry in interactions between large igneous provinces and mid ocean ridges, and their role in the kinematics of Shatsky Rise, an undersea volcanic plateau in the western Pacific Ocean that is roughly the size of the state of California. Shatsky Rise formed at a ridge-ridge-ridge (RRR) triple junction (TJ) (Hilde et al., 1976; Sager et al., 1988), at the intersection of the Pacific-Izanagi, Pacific-Farallon, and Izanagi-Farallon ridges, beginning at approximately 147 Ma. Investigations into Shatsky Rise’s formation are limited by the fact that two-thirds of the seafloor data has subducted with …

Investigation Of Laser Based Flow Diagnostics With Metastable Argon, Sterling S. Gordon

Physics Theses & Dissertations

The overarching motivation of this work is the development of seedless, non-intrusive, laser-based diagnostics for supersonic airflow in wind tunnels. To advance this goal, this dissertation fo-cuses on three-photon excitation in a research-grade argon beam within a tabletop vacuum system, which offers a controlled environment for testing and refining the approach. The chosen excitation scheme drives argon atoms to the 3d[5/2]₃ state using a pulsed Ti:Sapphire laser system, enabling time-of-flight (ToF) measurements on atoms that subsequently undergo a multi-step decay to the metastable 4s[3/2]₂ state. Although full realization of this excitation scheme was hindered by technical …

Piezoelectric Energy Harvesting From Roadways: Challenges, Advances, And Future Directions, Heba Gaber, Mohamed Abdelraheem

Faculty Publications

As the global demand for renewable energy intensifies, piezoelectric energy harvesting from roadways has emerged as a promising avenue for sustainable power generation. This systematic literature review analyzes 61 peer-reviewed studies to assess the feasibility, performance, and potential of integrating piezoelectric systems into roadway infrastructure. While technology faces challenges, such as high installation costs, limited energy output, and a scarcity of thorough economic evaluations, findings suggest it holds considerable promise as a supplementary renewable energy source. The review analyzes the operational characteristics and efficiencies of various piezoelectric transducers, identifies key factors influencing system performance, and evaluates recent technological advances. It …

Is Artificial Intelligence Really Taking Our Jobs?, Francesca A. Strom

Undergraduate Theses and Capstone Projects

This paper studies how artificial intelligence has reshaped occupational opportunities in the United States by analyzing changes in employment, median wages, and wage inequality across all occupations from 2014 through 2024. Using a difference-in-differences framework, I compare occupations associated with AI-related tasks to those not directly exposed to generative AI, with particular attention to the structural break introduced by the mainstream release of generative AI tools in 2022. The results show no statistically significant decline in employment among AI-exposed occupations, indicating that early adoption did not lead to measurable job displacement. Instead, the strongest effects appear in wage inequality. Among …

The Presentation Of Self In Everyday Digital Life: A Study Of Self- Disclosure And Work Environments, Mackenzie Michelle Skiff

Communication & Theatre Arts Theses

The digital age impacts individuals’ lives in many ways. One impact is how and where work is completed across many careers. The work-from-home strategy enables individuals to complete work that is not within a shared space, such as an office. With the absence of this shared space, communication practices within workplaces could be changing. Specifically, self-disclosure while working from home may differ from self-disclosure within the office or hybrid (both in office and remote) work environments. This thesis investigates whether there are differences in self-disclosure practices across three different types of contemporary work environments and offers a digital update to …

Making Explanations Make Sense: Xai For Smishing Detection, Eleni Alexandra Katsarakes

Psychology Theses & Dissertations

Explainable Artificial Intelligence (XAI) is a key component of effective human-AI collaboration, particularly in high-stakes domains such as cybersecurity. While AI tools hold promise for mitigating threats such as SMS-based phishing (SMiShing), their real-world effectiveness may hinge not just on detection accuracy, but on whether users can make sense of the system’s outputs. As SMiShing attacks grow in both frequency and sophistication, so does the urgency of designing human-centered AI systems that support user decision-making under uncertainty. This study examined how four distinct AI explanation types - Normative (rule-based), Attributive (feature-based), Exemplar (case-based), and Recommendation-Only - influence user performance, confidence, …

Component Model Development Of Heat Exchangers, Expanders, And Control Valves For Autonomous Cryogenics Plant Cool-Down, William Harris Buhrig Iv

Mechanical & Aerospace Engineering Theses & Dissertations

The traditional method of cryogenic plant cool-down involves having continuous on-call staff to head into the office at any time to modify the existing multi-layered PID control systems if the on-call staff member detects a significant deviation from the cool-down plan. This thesis aims to outline an effective method for modeling the structure of systems with performance characteristics that deviate from design requirements and from ideal inlet-outlet correspondence, enabling the adjustment and modification of existing control structures across all Thomas Jefferson National Accelerator Facility (JLab) cryogenic refrigeration plants. Analytical Modeling and Gaussian Process Regression (GPR) are applied to model the …

Design And Assessment Of A Single-Blade Rotary Wing For Use In Vertical Takeoff And Landing (Vtol) Aircraft, William C. Mcmasters

Mechanical & Aerospace Engineering Theses & Dissertations

A single-bladed propeller was designed to determine its efficacy in increasing efficiency of quadplane aircraft in the cruise configuration while still providing sufficient vertical lift for the vertical takeoff portion of flight. Aerodynamic theory predicts higher efficiency for single-blade propellers, resulting in a lower power requirement. A counter-weighted single-blade was modeled after a 10x5 model aircraft propeller used in small unmanned vehicles with a steel counterweight to balance centrifugal forces. The propeller was tested in the ODU wind tunnel to determine performance at various Advance Ratios (J). The research suggests that single-blade propellers show comparable performance compared to two-bladed propellers …

Sciteuq: Toward Uncertainty-Aware Complex Scientific Table Data Extraction And Understanding, Kehinde Ajayi

Computer Science Theses & Dissertations

Scientific tables report critical research insights, data, and findings for scientific progress. Because Portable Document Format (PDF) is the de facto standard format for scientific paper publishing, there has been an emerging need for an automatic method to extract data from PDF files. A significant fraction of scientific tables exhibit complex structure and content, making it challenging for machine learning tools to accurately extract the content directly from PDF files. Despite the advancements in Table Structure Recognition (TSR), automated extraction of data from complex scientific tables remains a challenge due to variations in table structures and contents. In this dissertation, …

Comparative Analysis Of Emergent Behaviors Of Three Drone Swarm System Models For Targeting Using Agent-Based Modeling And Simulation, Arsenio T. Gumahad Ii

Engineering Management & Systems Engineering Theses & Dissertations

This dissertation introduces a novel computational simulation framework for evaluating the emergent behaviors of three swarm drone models using Agent-Based Modeling and Simulation (ABMS). The three swarm models are a Leader-Follower swarm model based on Bruckstein's antline theory, a Flocking model based on a simplified Reynolds 'Boids’ model, and a Stigmergic model with pheromone-based coordination. The primary objective of the simulation is to evaluate the performance of these models in delivering a user-defined number of drones of each type to a target area of interest in four separate scenarios, resulting in 50,000 separate simulation trials. Each scenario was structured to …

Human Identification And Action Recognition Using Small Data And Deep Domain Adaptation, Alexander M. Glandon

Electrical & Computer Engineering Theses & Dissertations

Human identification and human action recognition problems are two important research areas for real-world security and surveillance applications. In both human identification and action recognition, it is necessary to operate by collecting small datasets in the field, possibly in a short time window of observation. This dissertation studies and develops computational modeling and high-performance machine learning (ML) and deep learning (DL) models for human identification and human action recognition using small amounts of data. These methods and computational models may be useful for different security and surveillance applications.

This dissertation on human recognition develops a ML computational model to estimate …

Margalefidinium Polykrikoides Group Iii Va, Usa Strain Growth And Sac-Like Pellicle Cyst Dynamics, Eduardo Pérez Vega

OES Theses and Dissertations

Margalefidinium polykrikoides is a harmful cosmopolitan dinoflagellate that blooms in coastal waters. The effect of temperature and salinity on the growth of M. polykrikoides VA strain was examined using microscopy and growth models. M. polykrikoides Group III VA strain grew better at warmer temperatures and lower salinities than M. polykrikoides Group III NY strain, Group I Korea strain, and Japan strain (unknown group). Modelers need to use the temperature and salinity growth responses from M. polykrikoides Group III VA strain to better simulate and predict M. polykrikoides blooms in the Chesapeake Bay.

Dinoflagellates produce cysts as a strategy to withstand …

Examining The Perceptions Of Economics And Personal Finance Teachers' Preparedness To Teach Financial Literacy, Tasha L. Wearren

Educational Leadership & Workforce Development Theses & Dissertations

Financial literacy has become an important topic in our society. Despite an increased focus in financial literacy education, there is limited research assessing teachers’ ability to teach financial literacy by examining their endorsement area. According to Soroko (2021), research is limited regarding teachers’ beliefs and practices as it relates to financial literacy. As O’Neill and Hensley (2016) indicated, “Teachers cannot teach personal finance well if they do not understand it themselves and/or cannot engage their students” (p. 639). This qualitative multiple case study examined how Economics and Personal Finance teachers perceive their preparedness to teach financial literacy with different endorsements. …

From Standard To Stratified: Modeling Ntcp And Ear To Personalize Daily Mv-Cbct In Radiotherapy., Duong Thanh Tai, Luong Tien Phat, Tran Trung Kien, Duong Tuan Linh, Nguyen Ngoc Anh, Nguyen Quang Hung, Peter Sandwall, Parham Alaei, David Bradley, James C L Chow

Oncology Articles

PURPOSE: To evaluate the cumulative radiobiological impact of daily megavoltage cone-beam computed tomography (MV-CBCT) imaging dose based on normal tissue complication probability (NTCP) and excess absolute risk (EAR) of secondary malignancies among radiotherapy patients treated for breast, pelvic, and head & neck cancers. This study investigated whether MV-CBCT imaging dose warrants protocol personalization according to patient age, anatomical treatment site, and organ-specific radiosensitivity.

METHODS: This retrospective study included cohorts of breast (n = 30), pelvic (n = 17), and head & neck (n = 20) cancer patients undergoing radiotherapy with daily MV-CBCT. Imaging dose distributions employing two common MV-CBCT protocols …

Smags-Lasso: A Novel Feature Selection Method For Sensitivity Maximization In Early Cancer Detection, Hamid Khoshfekr Rudsari, Sara Khorami-Sarvestani, Johannes F Fahrmann, James P Long, Samir Hanash, Kim-Anh Do, Ehsan Irajizad

Faculty, Staff and Student Publications

Background: Sensitivity and specificity are foundational metrics for cancer detection tools. However, most machine learning algorithms prioritize overall accuracy during optimization, which fails to align with clinical priorities of early detection. We aim to develop a feature selection machine learning algorithm while maximizing sensitivity at a given specificity.

Methods: We developed SMAGS-LASSO, a machine learning algorithm that combines our developed Sensitivity Maximization at a Given Specificity (SMAGS) framework with L1 regularization for feature selection. This approach simultaneously optimizes sensitivity at user-defined specificity thresholds while performing feature selection. SMAGS-LASSO utilizes a custom loss function with L1 regularization and multiple parallel optimization …

“Starlight Suite”: A Compositional Exploration Of The Argo Navis Constellations In Three Movements, Adam Miller

Electronic Theses and Dissertations

This thesis explores a compositional technique in which stellar constellations are arranged upon musical staves in a variety of iterations to create a basis from which to draw melodic, harmonic, and rhythmic material. These raw extrapolations of unaltered notes are then arranged and orchestrated into thematic music for full orchestra, centered around elements from the mythological background behind the constellations used. An accompanying analysis explains the specific applications of this technique within each movement of the work, the significance of each, and the variety of ways in which musical inspiration was drawn from the raw materials. The objective of this …

Knowledge Mapping And Visualized Analysis Of Research Progress In Onconephrology: A Bibliometric Analysis, Yiwei Wang, Shuling Fan, Wei Wang

Faculty, Staff and Student Publications

Objectives: Onconephrology is an expanding subspecialty focused on the management of cancer patients with renal injury. This study used a comprehensive bibliometric analysis to emphasize the need for cooperation between oncologists and nephrologists, exploring current trends and future research areas in onconephrology.

Methods: Relevant literature on onconephrology published between 1 January 2000 and 27 April 2024 was retrieved from the Science Citation Index Expanded of the Web of Science Core Collection, followed by manual screening. Bibliometric analyses were performed using CiteSpace, VOSviewer, and Bibliometrix software.

Results: A total of 1,853 publications, including 1,647 articles and 206 reviews, by 11,606 authors …

An Investigation Of Paleoclimate Through A Sedimentological Lens In The Gulf Of Mexico And Western North Pacific, Sarah Monica

Dissertations

This dissertation explores the use of sediment cores as proxy data for the reconstruction of paleoclimatic and environmental conditions. Human-caused climate change is leading to dramatic shifts in the global climate. As the instrumental record of climate is relatively short compared to the amount of time Earth has experienced weather, this work aims to extend our knowledge of climate beyond the instrumental record, thereby improving our holistic understanding of the climate system. In Chapter one, sediment cores from the central Texas inner shelf are used to produce a record of intense tropical cyclone (TC) activity over a ~4500-year period. X-Ray …

Molecular Design And Engineering Of Luminophores For Aggregation Induced Electrogenerated Chemiluminescence, Jesy Alka Motchaalangaram

Dissertations

Most conventional luminophores produce intense emissions in solutions but suffer from weak emissions or quenching when aggregated in poor solvents due to intermolecular interactions, such as π-π stacking. This phenomenon is known as aggregation caused by quenching (ACQ), limits their applications in their solid state. In contrast, aggregation induced emission (AIE) is a phenomenon in which luminophores are weak- or non-emissive in solution but emit intensively in their aggregated or solid states. AIE has enabled significant advancements in various real-world applications and has inspired new areas of research. The combination of AIE with electrogenerated chemiluminescence (ECL) has resulted in a …

Array Of Mini-Graphene-Silicon Solar Cells Intermittently Recharges Storage Capacitors Powering A Temperature Sensor, Ashaduzzaman, Syed M. Rahman, Md R. Kabir, James M. Mangum, Hung Do, Gordon Carichner, David Blaauw, Paul M. Thibado

Physics Faculty Publications and Presentations

Developing autonomous sensor systems that draw power from the ambient environment is a critical step for creating the Internet of Things. In this study, the authors built dozens of graphene-based solar cells, wire bonded them into standard packages, and characterized the current-voltage characteristics of each under illumination. Next, solar cells were connected in series to increase the output voltage. Three different sets of solar cells were used to charge three storage capacitors to the voltage levels required by our temperature sensor. The storage capacitors require only a few minutes to charge, yet power the sensor system for more than 24 …

Auction Consensus Algorithm With Loss Mechanism For Decentralized Task Allocation, Jose Rodriguez, Wenjie Dong, Constantine Tarawneh, Qi Lu

Electrical and Computer Engineering Faculty Publications

This paper presents an Auction-Consensus Algorithm with a Loss Mechanism (ACALM), a decentralized task allocation method for multi-robot systems that enhances the existing Consensus-Based Auction Algorithm (CBAA) by incorporating a novel loss propagation mechanism. In contrast to purely greedy bidding strategies, it enables agents to dynamically update task priorities based on the accumulated loss from previously unsuccessful bids. This extended work reduces globally inefficient allocations caused by early suboptimal decisions. The proposed approach is evaluated through large-scale simulations in thousands of randomized scenarios and swarm sizes ranging from 5 to 120 robots. Compared to existing CBAA and GCAA algorithms, ACALM …

Scalable And Durable Superhydrophobic-Superoleophilic Nanostructured Zinc-Coated Steel Wool For Efficient Oil-Water Separation, Nawzat Saadi, Laylan B. Hassan, Tansel Karabacak

Faculty Scholarship

We present a scalable, low-cost method to fabricate durable, superhydrophobic-superoleophilic steel wool for oil- water separation. The process involves zinc electroplating, hot water treatment (75◦C) to grow ZnO nanowires, and functionalization with 1-dodecanethiol. The resulting material exhibits a water contact angle of ~160◦ and an oil contact angle of ~0◦. It achieves a high oil absorption capacity (~18 times its weight) and separation efficiencies of ~99–100 % for cyclohexane (within 60 s) and crude oil (in under 3 min). The material demon- strates exceptional corrosion resistance in saline environments (14 wt% NaCl), retains its nanostructural integrity after prolonged exposure, and …

Zero Day Malware Detection With Alpha: Fast Dbi With Transformer Models For Real World Application, Matthew Gaber, Mohiuddin Ahmed, Helge Janicke

Research outputs 2022 to 2026

The effectiveness of an AI model in accurately classifying novel malware hinges on the quality of the features it is trained on, which in turn depends on the effectiveness of the analysis tool used. Peekaboo, a Dynamic Binary Instrumentation (DBI) tool, defeats malware evasion techniques to capture authentic behavior at the Assembly (ASM) instruction level. This behavior exhibits patterns consistent with Zipf's law, a distribution commonly seen in natural languages, making Transformer models particularly effective for binary classification tasks. We introduce Alpha, a framework for zero-day malware detection that leverages Transformer models, Support Vector Machines (SVMs) and ASM language features. …