Digital Commons Network^™

Microplastics In St. John, United States Virgin Islands Sediments: Methodological Approaches And Variability Of Microplastics Concentrations And Accumulation Rates, Henry B. Arbaugh, Sarah C. Gray, Dimitri Deheyn

Theses

Sediment samples were collected from a near shore and reef site in Coral Bay, St. John, United States Virgin Islands (USVI) from 2007-2016, using tube sediment traps. These samples provide an opportunity to study microplastics (MPs) in an area where no previous MP studies have been performed. Most sedimentary MP studies utilize benthic samples which can only measure MP concentrations (#MP/g or #MP/mL sediment), while sediment traps can provide MP accumulation rates (#MP/cm²/day) as a function of time. However, the methods used for extracting MPs from benthic samples have not been tested using sediment trap samples whose characteristics, …

Spike-Based Architectures For Energy-Efficient Audio Enhancement, Kedar Sandip Fitwe

Theses

Deep Neural Networks (DNNs) have shown remarkable success in speech denoising; however, their high computational and energy requirements make realtime deployment on edge devices challenging. In contrast, Spiking Neural Networks (SNNs) operate using sparse, event-driven spikes—offering a biologically inspired and energy-efficient alternative. In this study, we delve into a Spiking Neural Network (SNN) model that leverages the temporal dynamics of spiking neurons to capture long-range dependencies in the audio signal. By encoding the input audio into sparse and event-driven computations, the SNN can efficiently process the temporal information while requiring significantly fewer computations compared to DNNs. We present a real-time …

Evaluating The Effectiveness Of Smart City & Iot Solutions In Improving Urban Sustainability, Mobility, And Quality Of Life In The Uae, Aisha Alnuaimi

Theses

The United Arab Emirates has an aim to develop smart cities incorporating advanced technologies such as the Internet of Things (IoT), to enhance urban sustainability, mobility, and improve citizens quality of life. However, based on the significant funds invested by the government, there exists an important gap within the government systems in measuring the effectiveness of such activities given the lack of data quantifying the benefits gained from such smart initiatives. To address this gap, this study delivers the benefits gained from smart initiatives by using a mixed-methods approach combining both quantitative and qualitative data. The research is based on …

Investigating The Knowledge, Attitudes And Practices Towards Type 2 Diabetes Mellitus In A Student Population In The United Kingdom: A Quantitative Study, Philip Boadu Anim

Theses

Aim: To explore the knowledge on the risk factors, signs and symptoms, management, and complications of Diabetes, attitudes, and practices towards prevention of Diabetes Mellitus among university students in the United Kingdom.

Methods: The researcher utilised a quantitative study design in carrying out the study. The study was conducted at Cardiff Metropolitan University with Cardiff Met students being the target population. Thirty (30) participants were recruited through convenient sampling. Data was collected through questionnaires using closed ended questions. The Statistical Package for Social Science (SPSS) version 27 as used to analyse the data collected. Statics were performed to form a …

Assessing Localized Recovery And Synchrony Of An Endangered Wetland Species., Aiyana Reissman

Theses

The state and federally endangered light-footed Ridgway’s rail (LFRR; Rallus obsoletus levipes) is central to regional conservation, yet the drivers of its population fluctuations remain poorly understood. The fragmented nature of coastal wetlands in Southern California is threatening the existence of endemic species like the LFRR. Tidal inundation, severe storms, and overpredation threaten the survival of LFRR. This study addresses population fluctuations by analyzing the spatial and temporal drivers of LFRR populations to inform effective, site-specific management.

To do this, a 45-year (1980–2024) LFRR census data set across 16 wetlands throughout its range in California was synthesized. A multi-faceted …

Multipath Exploitation For Radar Sensing In An Urban Environment Using The Music Algorithm For Jammer Mitigation, Daniel De Zutter

Theses

With the increased push for spectrum sharing techniques, the technological world surrounding radar has been constantly pushing to find new methodologies. From being able to integrate with a communication system to mitigating the effects of other users on a system, the research community is pushing forward. We look to tackle the problem of mitigating a jammer using a multipath MUltiple SIgnal Classification (MUSIC) based approach. We take multiple steps to implement this and generate a framework for future work. We first take a look at understanding the current state of the art research surrounding DFRC, multipath exploitation, MUSIC algorithms, RIS …

Knowledge And Perception Of Human Milk Donation Among Pregnant, Lactating, And Postpartum Women In Rochester, Ny, And Near Hmbana Milk Banks, Samantha Allen

Theses

Human milk banks provide pasteurized donor milk for premature infants whose mothers are unable to produce sufficient amounts of milk, offering critical nutritional and immunological benefits for this high-risk population.1–4 Although 28 HMBANA-accredited milk banks operate in the United States, none are located within 4.5 hours of Rochester, NY. This study examined the knowledge and perceptions of human milk bank donation among pregnant, lactating, and postpartum women in Rochester, NY, and in communities near HMBANA-affiliated milk banks, and assessed if there is a correlation regarding the proximity of milk banks.  A cross-sectional study design was used, and participants completed a …

Customer Lifetime Value Prediction In E-Commerce, Shaimaa Alantali

Theses

The e-commerce market across the world currently deals with an urgent problem which has emerged because businesses now use automated pricing systems everywhere. Online merchants who automate their decision processes to handle changing customer needs and limited stock situations must face external business forces which push their teams to start destructive pricing battles that harm all businesses in the sector. The research investigates how a unified computational framework which combines demand modeling with advanced data analytics methods can enhance pricing strategies through market instability detection and prevention systems. The study creates a market simulator through data-driven methods which uses retail …

Transformation Of Public Earth Datasets For 3d Narrative Visualization Of Disaster Events, Connor Bleisch

Theses

Communicating large datasets in ways that are appealing to non-scientific audiences has been a recent focus of big data visualization. Interest has grown for using 3D environments with game-like structures and immersive technology like AR and VR since they are more engaging. We introduce a method, based on the structure of the datasets and domain knowledge of the natural phenomena, for transforming public National Aeronautics and Space Administration (NASA) Earth datasets into 3D narrative data visualizations, specifically of disaster events. Two visualizations were created using this method: one of the KNP Complex Wildfire in 2021 and one of Hurricane Beryl …

Nandworks : A Low-Cost Testbed For Interfacing Onfi-Compatible Nand Flash Memory Devices, Mark Scheidker

Theses

NAND flash memory is the basis for modern digital storage, but direct access to raw flash functionality remains difficult due to limitations in available interface tools. Industrial characterization systems can issue low-level commands but are prohibitively expensive for most researchers and students. USB-based bit-banging tools lack the timing fidelity needed for certain experiments. This thesis introduces NANDWorks, a testbed based on the Raspberry Pi 4 that provides direct access to ONFI-compatible NAND flash devices via GPIO bit-banging in Linux user space, building on prior work for the Terasic DE1-SoC platform. NANDWorks has a layered software stack that includes a hardware …

Uae Traditional Folklore Through Modern Media And Its Educational Potential: The Case Of Freej, Shahad Imad Mostafa

Theses

This mixed‑methods study examined how the Emirati animated series Freej preserves traditional folklore through vernacular language, dialect, and idiomatic expressions, and evaluated its educational potential. Guided by four questions on folklore representation, dialect/idiom use, the role of the characters, and curricular applicability, the research combined content analysis of selected episodes, interviews coding and a survey of 100 Emirati participants; survey data were summarized with descriptive statistics to triangulate qualitative insights.

Findings indicated that audiences overwhelmingly perceived Freej as linguistically and culturally authentic: 95% rated the dialect as very/somewhat authentic, 94% said it reflects elders’ speech, and 89% felt the series …

Identification And Analysis Of Transverse Cirrus Bands In Tropical Cyclones Using A U-Net Model, John Mark Mayhall

Theses

Transverse cirrus bands (TCBs) are bands of upper-level clouds commonly observed in mesoscale and synoptic-scale weather systems. In tropical cyclones (TCs), their appearance has been hypothesized to relate to TC intensity, intensification, and the diurnal cycle. Rigorous testing of these relationships has been limited by the subjective nature of identifying TCBs in satellite imagery. This paper describes a convolutional neural network (CNN) that successfully identifies TCBs in satellite imagery. Using the CNN, a climatology of TCBs from 2019 to 2023 was developed for the Atlantic and Eastern Pacific TC basins. The climatology indicates that TCBs display a diurnal cycle with …

Impact Of Oxidative Stress On Mapk Signaling In Human Pancreatic Ductal Cell Line, Alexis Fink

Theses

The main processes linked to diabetes (type 1 or type 2) and obesity are impaired insulin production and glucose intolerance. The imbalance between antioxidant defense systems and oxygen-reactive species are known as oxidative stress. Understanding oxidative stress and its substantial effects on diabetes is crucial. Reactive oxygen species (ROS) can activate mitogen-activated protein kinases (MAPKs). We hypothesize that altered MAPK phosphorylation and hence downstream signaling is associated with changes in oxidative stress in the human pancreatic duct epithelial cell line (H6C7). While H6C7 cells are often engineered to be cancerous precursors, the natural phenotype of these cells and their expression …

Exploring Interpenetrating Networks Derived From Compostable Polymers In Separations, Hailey Holshouser

Theses

The increasing demand for environmentally friendly and sustainable water treatment technologies drives the development of membranes that work to balance mechanical durability, tunable selectivity, and long-term stability. In this exploratory work, an interpenetrating polymer network (IPN) membrane system composed of poly(butylene adipate-co-terephthalate) (PBAT) and poly(2-hydroxyethyl methacrylate) (pHEMA) was developed to provide a more compostable and compatible platform for advanced separation applications. While the compostability of the entire framework was not tested, the dominant presence of PBAT in the matrix makes the membrane intrinsically favorable for composting. Mechanochemical Ionic Electrochromic Generators (IEGens) were synthesized and incorporated into the PBAT backbone for …

Experimental Characterization Of Gas-Injected Bubble Dynamics In Various Fluids And Injector Arrays For Centrifugal Nuclear Propulsion Applications, Julia Kondrat'yev

Theses

This study experimentally characterizes gas-injected bubble dynamics in analog fluids to establish baseline behavior for Centrifugal Nuclear Thermal Propulsion (CNTP) systems. Nitrogen was injected through mass flow ranges of 3.3E-05 to 9.6E-05 lbm/s through various injector arrays in a transparent tank apparatus using water or sodium polytungstate (SPT-3) to vary fluid properties. Four injector configuration cases were tested on a 4-in deep, clear-sided chamber at atmospheric surface pressure to examine the effects of fluid viscosity and density, circular injector diameter, injector location, and mass flow rate on bubble diameter, rise velocity, and plume structure. Bubble equivalent diameters, positions, and velocities …

Urban Effects On Cloud Formation, Boundary Layer Circulation, And Rainfall Over Miami, Aaron A. Serre

Theses

The sensitivity of cloud and precipitation formation is expected to be enhanced over urban areas due to persistent sea-breeze forcing. This study combines surface observations, satellite data, and numerical modeling to investigate the impacts of urbanization in the Miami Metropolitan Region (MMR) on surface meteorology, cloud formation, and precipitation. Satellite observations reveal urban expansion of approximately 348 km², along the western boundary of the MMR. These areas also exhibit statistically significant increases in land surface and air temperatures. Geostationary satellite data show enhanced morning cloud formation over the MMR, while increasing rainfall and interannual variability in cloudiness are observed along …

Effects Of Energetic State On Male Mating Preferences In Poeciliid Fishes With Differing Mating Strategies, Katherine E. Taylor

Theses

Reproduction is energetically costly, making condition (i.e., energetic state) a potential factor in shaping reproductive behavior. However, its role in male mate choice remains largely unexplored. Here, we investigate condition-dependent effects on male mate choice in two species of Poeciliids: one coercive species (Gambusia affinis) and one courting species (Xiphophorus hellerii). Diet treatments were used to manipulate male and female conditions, and male preferences were assessed in dichotomous-choice tests using animated female stimuli. X. hellerii males showed a preference for high-condition females, whereas G. affinis males showed no overall net preference. Diet influenced preference in G. affinis, with high-fed males …

Inflex : A Scalable Framework For Malware Analysis, Correlation, And Steganography Detection, Luke Bower

Theses

Modern malware campaigns operate in stages across diverse formats, including PE, ELF, shellcode, and malicious documents, some leveraging steganography to deliver payloads such as Quasar RAT and Lokibot. Because analysis tooling remains siloed by artifact type, analysts struggle to identify shared code reuse, behavioral overlap, and campaign relationships when adversaries avoid traditional indicators. This thesis presents INFLEX, a framework that normalizes diverse artifacts through a unified pipeline combining static analysis, emulated behavioral extraction, sandbox execution, and threat intelligence enrichment. Its correlation layer introduces Variant Resilient Function Hashing (VRFH), validated against fuzzy hashing baselines on targeted samples, and Weighted Artifact Fusion …

The Influence Of Defendant Mental Illness Characteristics On Prosecutorial Decision Making, Claire Elise Martin

Theses

As most criminal cases are now resolved via plea-bargaining, research on legal decision making has extended beyond the jury to evaluate judges, defense attorneys, and prosecutors. Although legal actors are obliged to remain objective and dispassionate, extralegal factors regarding defendant characteristics (e.g., age, gender, race/ethnicity, socioeconomic class) can influence case outcomes. Extant research on the effect of defendant mental health status and defendant demeanor on sentencing outcomes in court hearings is limited. Therefore, the present experiment solicited 134 prosecutors in a vignette study designed to assess how extralegal characteristics, specifically those related to mental illness, may affect the punitiveness of …

Modeling And Identification Of Thermal Interactions In Lithium-Ion Battery Modules, Saima Alam

Theses

The recent development of lithium-ion battery technology has fueled its adoption in electric vehicles (EVs) as the primary power source due to its high energy and power density. However, the high energy density comes at the cost of higher safety requirements. To ensure the lithium-ion battery pack's operational safety, performance optimization, and longevity, a Battery Management System (BMS), an electronic controller, is employed. Recent reports of battery pack overheating, fire, and explosion in EVs expose the limitations of the BMS in detecting the abnormalities and preventing them from propagating to other cells. To address these limitations, in the first part …

Development And Validation Of An Automated Image-Based Framework For Bubble Size And Velocity Measurement In Centrifugal Nuclear Thermal Rocket Experiments, Olivia Williams

Theses

Accurate measurement of bubble size and rise velocity is essential for characterizing multiphase flow behavior in Centrifugal Nuclear Thermal Rocket (CNTR) experiments, where dynamic imaging conditions complicate traditional analysis methods. This thesis presents the development and validation of an automated image-based framework for extracting bubble equivalent diameter and rise velocity from high-speed experimental videos of injected gas bubbles rising through controlled liquid-filled volumes. Representative datasets include air bubbles injected into water and Sodium Polytungstate (SPT-3). The framework employs a staged analysis workflow with instance-level bubble segmentation using a Mask R-CNN model, secondary classification for detection verification, multi-object tracking, and velocity …

High Strain Rate Behavior Of Soft Polymeric Materials : Polymer-Bonded Explosive Simulants And Polyimide Aerogels, Benjamin Jeffrey Solomon

Theses

Soft polymeric materials exhibit strain sensitive behaviors and are used in various applications where dynamic loading conditions may be present, making understanding the responses at high rates vital to safe application. Investigation of the mechanical response of soft materials at high strain rates has proven challenging due to low forces, mismatching impedances, and longer times to obtain dynamic force equilibrium limiting the quality and accuracy of data acquisition. Two polymeric materials were evaluated in this study. The first is polymer-bonded explosive (PBX) 9501, which experiences dynamic loading during use as solid propellant and warhead explosive material. It is found that …

Discovering The Genes And Molecular Mechanisms Involved In Increasing The Expression Of The Gad Regulon In Escherichia Coli Resistant To Macrolide Antibiotics, Lea V. Freeman

Theses

Various genes have been proposed to play a role in the action of indole, a bacterial hormone involved in biofilm and quorum sensing. Previous work found that E. coli strains carrying a macrolide-resistant mutation in the uL22 ribosomal protein reduce tna operon mRNA levels and, consequently, decrease indole production. This ribosomal mutation also increases expression of the gad regulon, a genetic unit involved in acid resistance at pH below 2. The gad regulon is involved in bacterial survival under environmental challenges, primarily regulating intracellular acid resistance, and its expression appears to depend on indole production. In this work, we work …

Optical Turbulence Estimation From Numerical Weather Prediction Models: Methods And Error Analysis, Richard Minor

Theses

Optical turbulence plays a critical role in electromagnetic wave propagation through the atmosphere, directly affecting imaging systems, wireless communications, and high-energy laser (HEL) performance. Realistic three-dimensional characterization of atmospheric turbulence is essential for understanding these impacts, particularly within the atmospheric surface layer, where turbulence is often strongest. The refractive index structure parameter, Cn2, is used to quantify optical turbulence. This thesis evaluates the capability of the High Resolution Rapid Refresh (HRRR) model to characterize surface layer Cn2. As the highest resolution and most frequently updated numerical weather prediction model covering the continental US, HRRR offers strong potential for operational HEL …

What Correctness And Faithfulness Cannot Tell You : Counterfactual Behavioral Diagnosis Of Rag Systems, Saif Raza Khan

Theses

Correctness and faithfulness are standard metrics used for evaluating Retrieval-Augmented Generation (RAG) systems, but they only describe the outcomes and do not surface the underlying behaviors behind them. The same scores can hide a number of underlying behaviors. They can hide whether a system relied on retrieval, answered from parametric knowledge, followed a near-miss context, missed the gold source or failed to use the retrieved evidence. The same scores can hide all of these. This thesis introduces a counterfactual behavioral evaluation framework that tests each query under five controlled context conditions and uses a gold-source coverage gate to identify whether …

Drift Aware Continual Finetuning Of Encoder-Only Text Classifiers Under Temporal Distribution Shift, Simran Kc

Theses

The performance of encoder-only text classifiers degrades silently as their input distribution drifts over time. This thesis presents a closed-loop pipeline for drift-aware continual fine-tuning that combines a sample size invariant drift severity metric, an automated retraining loop on the most drifted samples, and a validation gate that rejects unsafe updates before deployment. We evaluate the pipeline on two datasets, Amazon Electronics Reviews and HuffPost News Category, and observe significant performance improvements over the initially trained model, along with a substantial reduction in compute and labelling costs compared to fresh-per-year and cumulative retraining baselines.

Performance Comparison Of A 25 Mm Rotating Detonation Rocket Engine Using Multi-Ring Injection, Nathan C. Mead

Theses

Rotating detonation rocket engines (RDREs) use one or more detonation waves traveling circumferentially around an annular (center body, CB) or cylindrical (center bodiless, CBL) combustion chamber, offering theoretical advantages over conventional deflagration-based combustors. A new geometry, the center fill (CF), further enhances the CBL geometry by adding a fuel-oxidizer injector ring near the center of the combustion chamber, increasing energy density and thus propulsion performance. To evaluate CF operability and performance, a 25 mm CF RDRE running gaseous methane and oxygen is tested across varying equivalence ratios at 0.076 kg/s total mass flow rate, and across varying mass flow rates …

Impacts Of Vertical Wind Shear And Coastal Friction On The Pre-Landfall Rapid Intensification Of Hurricane Nicholas (2021), Ethan O. Prunty

Theses

Predicting tropical cyclone (TC) rapid intensification (RI) under moderate-to-strong vertical wind shear (VWS) remains challenging, particularly when cyclogenesis and RI occur within 48 hours of landfall, as with Hurricane Nicholas (2021). This study analyzes Nicholas’ RI using NOAA’s Hurricane Analysis and Forecast System. Results reveal a two-stage vortex alignment before the early RI stage: diabatic heating first draws the mid-level center toward the low-level center, followed by cyclonic precession. A nearly-saturated core below 500 mb develops before RI onset as column-integrated moist static energy (MSE) increases, and the moistening process continues during the early RI stage as surface heat and …

Deep Joint Source-Channel Coding For Semantic-Aware Adaptive Wireless Image Transmission, Sudipto Das Suvro

Theses

The growing use of AI-enabled applications has increased the demand for adaptive, efficient, and low-latency image transmission of large data volumes. This has created challenges for conventional separate source-channel coding that focuses on reliable bit sequence transmission without considering the contents or communication objectives. Deep joint source-channel coding (DeepJSCC) potentially solves this by combining source compression and error correction into a unified process. This thesis focuses on developing adaptive and computationally lightweight DeepJSCC-based transceivers for image transmission across heterogeneous channels. The framework further investigates DeepJSCC’s performance in OFDM with robust channel estimation and importance-based subcarrier allocation strategies. Comprehensive evaluation using …

Evaluating Datasets And Models For Offensive Language Identification, Skye Deson Morgan

Theses

Social media platforms enable people to easily communicate with one another, share ideas and express their opinions. Unfortunately, the anonymity provided by platforms such as Twitter and Facebook is more than enough to foster the spread of harmful content. Government agencies, as well as technology companies, are actively working towards finding better means of curbing offensive and objectionable content in social media and contributing to a more welcoming online environment. In this study, I first present a critical literature review on offensive language identification, exploring offensive language datasets and computational models using both machine learning and deep learning. Next, I …