Digital Commons Network^™

Understanding Machine Learning Model Behavior Under Fairness And Privacy Constraints, David Anthony Sanchez

Open Access Theses & Dissertations

Machine learning systems deployed in high-stakes domains are increasingly expected to satisfy demands beyond predictive accuracy-including fairness across demographic groups, protection of sensitive information, and explanations that human stakeholders can inspect and trust. This thesis investigates how those demands can be met through learning frameworks that explicitly govern the relationship between data and models, arguing that trustworthiness is a design problem rather than a post hoc correction. The thesis is organized around three studies, each targeting a distinct point of data-facing control. The first develops CondFairGen, a fairness-aware conditional generator for tabular data that improves subgroup equity by dynamically reweighting …

A Numerical Method For The Phase Field Crystal Model, Patrick Ameyaw Tabiri

Open Access Theses & Dissertations

The Phase Field Crystal (PFC) model is a continuum-based framework used to study theevolution of crystalline materials while preserving microscopic structural features over diffusive time scales. It captures complex phenomena such as phase transitions, defect dynamics, and microstructure formation via a nonlinear sixth-order partial differential equation derived from a free-energy functional. In this context, the phrase "sixth-order" indicates that the highest spatial derivative appearing in the equation is of order six. In this work, we study the mathematical formulation and numerical approximation of solutions to the PFC model. Due to the high-order spatial derivatives and nonlinearity in the governing partial …

Soil Microbial Responses To Artifical Light At Night, Randall L. Walker

Open Access Theses & Dissertations

Global change, driven by both natural processes and human activities, has significantly transformed ecosystems. Urbanization and artificial light at night (ALAN) are key contributors, disrupting natural light cycles and influencing biological rhythms, biodiversity, and ecosystem functions. Thus far, there remains a gap in knowledge of how soil ecosystems are impacted by ALAN. This study uses a laboratory incubation to investigate the effects of ALAN on soil microbial communities and functions across urban and natural ecosystems. Briefly, soil was collected from eight sites around El Paso, Texas and neighboring New Mexico representing urban and rural light exposure. Soil was then subjected …

Riemannian Manifold-Coupled Joint Sparse Graphical Models For Spatially Structured Precision Matrix Estimation, Abdul Rahman Adam

Open Access Theses & Dissertations

This thesis introduces a novel geometry-aware framework for joint precision matrix estimation in spatially structured data. Unlike existing joint Gaussian graphical model estimators that rely on Euclidean coupling mechanisms such as fused or group graphical lasso, our approach incorporates intrinsic Riemannian geometry through geodesic distance penalties on the symmetric positive definite manifold.

The method measures similarity between precision matrices of adjacent spatial regions using the affine-invariant Riemannian distance, capturing deformations of concentration ellipsoids rather than entrywise differences. This is particularly suited for spatial transcriptomics, where regions may share dependence structures despite differences in scale or composition.

An efficient Alternating Direction …

Exhumation Of The Pelona-Orocopia-Rand Schist And Fault Kinematic Constraints On Evolution Of The Eastern California Shear Zone, James Boyette Barker

Open Access Theses & Dissertations

Southeastern California has experienced a complex recent geologic history as it continues to evolve from a convergent boundary to a period of NE-directed extension and evolution of the Eastern California Shear Zone (ECSZ). This transition is recorded in exhumed bodies of Pelona-Orocopia-Rand (POR) schist and in changing styles of fault deformation. This thesis has two interrelated goals. First, I aim to refine the timing of transition from NE extension to NS contraction in the ECSZ of southeastern California by integrating (U-Th)/He thermochronology with field-based structural analysis. The second prong of this study is to constrain broader patterns of POR schist …

Bayesian Deep Learning For Photovoltaic Power Forecasting: A Probabilistic Framework For Uncertainty Quantification And Grid Reliability Optimization, Pablo Abraham Bustamante

Open Access Theses & Dissertations

The global energy landscape is undergoing a profound transformation, driven by the urgent need to decarbonize power systems, enhance energy security, and meet growing electricity demands. Solar photovoltaic (PV) power has emerged as a critical component of future energy infrastructure due to its abundance, scalability, and cost-effectiveness. However, PV generation is inherently variable and weather-dependent, introducing significant uncertainty into grid operations and complicating the task of balancing supply and demand. Accurate forecasting of PV power generation-particularly on day-ahead and hour-ahead horizons-has become a strategic necessity for grid stability, economic efficiency, and environmental sustainability. PV output is influenced by numerous factors, …

From Infiltration Pathways To Aquifer Recharge: An Electrical Resistivity And Vadose-Zone Isotope Study In Pecan Agroforestry Systems, Jose Josue Cabral

Open Access Theses & Dissertations

Groundwater in arid or semi-arid regions is critical for agricultural production, yet the contribution of irrigation return flow to regional aquifer recharge remains poorly constrained. As is the case beneath flood-irrigated pecan orchards in the Rio Grande floodplain of Far West Texas. This study integrates time-lapse electrical resistivity tomography (ERT) with stable water isotopes (δ¹⁸O, δ²H) in vadose-zone soil-pore water to evaluate how irrigation water infiltrates. This study tracks how irrigation return is stored and may contribute to recharge of the Hueco Bolson aquifer. The objectives are to (1) relate observed resistivity variations to stratigraphic heterogeneity influencing infiltration, (2) characterize …

Exploring Connectivity And Element Distribution In The Critical Zone Using Deep Cores Below Agricultural And Natural Drylands In The American Southwest, Angie Lynette Cano

Open Access Theses & Dissertations

Water movement controls how solutes are transported and stored in arid Critical Zone (CZ) systems. In natural drylands, low precipitation and high evapotranspiration lead to long term salt accumulation in the CZ, whereas irrigation increases water input and increases downward solute transport. This study investigates how agricultural practices alter subsurface hydrology and geochemistry by comparing two deep CZ cores from the southwestern United States: a 50 m deep core beneath a flood irrigated pecan farm named 5R farm, in the Rio Grande Valley in Tornillo, Texas, and a 97.5 m deep core from the Red Lake Playa in the Jornada …

Timing And Synchronization: Software And Hardware Approaches For Precision Maximizing, Christina Alicia Alejandra Carreon Jimenez

Open Access Theses & Dissertations

This thesis explores the design of synchronization mechanisms that allow multiple FPGA boards to operate as if they were a single coherent timing domain. The solutions span three layers: a software PLL for E1 telecommunication links, a PCIe-based timestamp exchange protocol, and a hardware-level multi-FPGA mesh architecture. Across these layers, the work demonstrates that practical, low-cost synchronization can achieve sub-microsecond accuracy without the burden of specialized hardware such as GPSDOs or White Rabbit infrastructure. Experiments show stable alignment within the nanosecond-microsecond range, validating the feasibility of the proposed methods for SDR and telecom applications.

Geothermal Gradient And Lithium Co-Generation Power Potential In Lafayette County, Arkansas, Juan Agustin Falquez

Graduate Theses and Dissertations

The Smackover formation in south Arkansas is a known reservoir for brines with high temperature and salinity. Recently, Smackover formation brines in Lafayette County, Arkansas are a target for lithium extraction. A county-wide dataset of Bottomhole Temperatures (BHT) is used to estimate a lower bound geothermal gradient of 36 C/km and a reservoir temperature of 116 C at an approximate depth of 2725 m. The high temperatures make the brines a candidate for geothermal power co-generation. Geothermal power co-generation can be achieved by implementing an Organic Rankine Cycle (ORC) power plant. The Southwest Arkansas Lithium project anticipates a total of …

Modeling Causal Interactions Across Brain Functional Systems For Population-Specific Disease Analysis, Alissen Moreno

Theses and Dissertations

Functional brain connectivity provides critical insight into neural mechanisms underlying neurodegenerative and affective disorders. Traditional neuroimaging studies rely on undirected or region-specific connectivity measures developed predominantly using racially homogeneous cohorts, limiting fairness and generalizability across diverse populations.

We propose a population-aware framework modeling directed causal interactions across large-scale brain functional subnetworks. Resting-state fMRI data from the HABS-HD cohort were used to construct subject-level causal connectivity matrices via ICA-LiNGAM, aggregated into interpretable system-level hyper-connectomes representing interactions among eleven canonical brain subsystems. These features trained nonlinear models for Alzheimer’s disease stage classification and trait worry prediction.

Results demonstrate that MLP models outperform …

Agentcite: Trustworthy Ai Agents For Information Verification Across Referenced Documents, Eymen Yigit

Master's Theses

This thesis presents AgentCite, a multi-agent framework for automated verification of referenced numerical data in research documents. The framework employs a hybrid design in which LLM-based agents handle document understanding and evidence retrieval, while deterministic components manage structured parsing and value comparison, improving consistency and reproducibility while reducing token consumption and execution time compared to fully agentic approaches.

AgentCite consists of three autonomous agents: a Negotiator, a Main Document Agent, and a Source Documents Agent, coordinated through a fixed tool-call pipeline. The Negotiator orchestrates verification by extracting tabular data from the main document, retrieving evidence from per-source vector stores, and …

Probing Representational Emergence In Large Language Models, Shawn Ismail

Master's Theses

This thesis investigates whether abrupt behavioral gains in large language models under scaling are accompanied by systematic changes in internal representations. It combines a behavioral screen of 65 tasks per family with targeted layerwise probing across eight decoder-only, open-weight model families. Behavioral emergence is defined for each family-task trajectory using an empirical jump detector, with segmented regression retained only as a diagnostic. The representational follow-up analyzes 27 selected MMLU subtasks shared across all families, spanning 37 checkpoints and 216 family-task units.

For each follow-up checkpoint, frozen linear probes are trained on every layer's hidden states to measure how much task-relevant …

Fret Studies On The Binding Of Fibroblast Growth Factor (Fgf) Proteins To The Hard Corona Of Cuins2-Based Quantum Dots, Colette Robinson

Graduate Theses and Dissertations

Fibroblast Growth Factor (FGF) has been shown to be an important protein in angiogenesis, which is critical in developmental biology and wound healing but has also been implicated in cancer metastasis. FGF interacts with an FGF Receptor (FGFR) to signal the onset of angiogenesis but the detailed mechanism of angiogenesis, and particularly its regulation, is still largely unknown. Imaging the interaction of FGF with FGFR in biologically-relevant environments, such as directly in cells, is key towards furthering this understanding and may eventually aid in the development of better cancer treatments. CuInS2/ZnS quantum dots (QDs) are an attractive fluorescent probe to …

High Throughput Phenomics Pipeline For Pulse Crop Nutritional Breeding, Amod Udayanga Madurapperumage

All Dissertations

Dry pea (Pisum sativum L.), lentil (Lens culinaris Medik.), and chickpea (Cicer arietinum L.) are major pulse crops valued for their high nutritional composition and importance to global food systems. Pulses are rich in carbohydrates, protein, and essential minerals, making them ideal whole foods and critical contributors to food and nutrition security. Due to these advantages, pulse breeding programs are increasingly focusing on enhancing nutritional traits, such as protein quality, amino acid balance, and micronutrient density, through the process of biofortification. However, improvement of agronomic traits remains equally essential. Characteristics such as plant height, standability, stress tolerance, …

Towards Generalizable Representation Learning Across Domains, Hossein Kashiani

All Dissertations

Despite remarkable progress in deep learning, a major challenge remains: machine learning models often struggle to generalize to unseen domains under distribution shift. In real-world settings, data often differ from training conditions due to changes in lighting, sensor type, image resolution, and style. These differences can significantly degrade performance, highlighting the need for representations that are both robust and generalizable. This thesis addresses this challenge by developing a set of frameworks for generalization across domains in anomaly detection, deepfake detection, and vision-language image recognition. For anomaly detection, this thesis introduces ROADS, a robust prompt-driven framework for multi-class unified anomaly detection. …

X-Raying The Material Surrounding Active Supermassive Black Holes, Isaiah S. Cox

All Dissertations

It is important to understand the nature of the gas and dust immediately surrounding the supermassive black hole (SMBH) in active galactic nuclei (AGN), because this material feeds the accretion disk and couples the activities going on at the AGN scale to the host galaxy. However, there remain large uncertainties on basic properties about this material such as its location and distribution. This is primarily due to the difficulties in resolving spatial information on these scales at extragalactic distances.

AGN emit a tremendous amount of radiation all across the electromagnetic spectrum, including X-rays. The X-rays come from a very small …

Geography, Income, And The Incidence Of A Massachusetts Carbon Tax, Susan Stratton Sayre

Economics: Faculty Publications

I use spatial microsimulation to estimate the geographic incidence of the consumer-facing portion of various revenue-neutral carbon tax designs in Massachusetts. I identify a substantial urban-rural disparity in impacts that is reduced, but not eliminated, by proposed revenue-targeting mechanisms. I further demonstrate that seemingly small variations in the definition of rural households can have substantial impacts on urban-rural equity and the distribution of gains and losses. Proposed policies are highly progressive in the aggregate with most low-income households receiving more in rebates than they pay in taxes, but notable numbers of low-income households are likely to experience significant losses. Methodologically, …

Conditioning Hierarchical Diffusion Transformers On Rf Circuit Parameters, Ethan Morton

McKelvey School of Engineering Graduate Student Theses & Dissertations

As modern ML techniques have become increasingly advanced, they have begun to be integrated into wireless RF systems for classification, identification, and spectrum management. Deep Neural Networks (DNNs) enable RF system operators and designers to design more flexible systems with greater robustness to errors and attacks. However, neural networks require significant amounts of properly annotated data to train. Current data labeling methods lack the ability to obtain reliable true labels for circuit properties such as carrier frequency offset (CFO), power amplifier (PA) non-linearity, and in-phase/quadrature (IQ) imbalance. This thesis investigates the efficacy of a novel architecture, RF-Diffusion, for generating high-quality …

Modeling Water Quality And Cyanobacteria Blooms In St. Lucie Estuary, Florida, Mohmmad Masudur Rahman

Electronic Theses and Dissertations 2020 - Present

The St. Lucie Estuary (SLE) on Florida’s east coast is vital for diverse aquatic life and provides essential services to local communities. However, the estuary’s health is threatened by water quality deterioration due to nutrient-driven phytoplankton blooms, particularly cyanobacteria, which have increased in frequency over recent decades. These blooms may worsen water quality by reducing dissolved oxygen, altering pH, and increasing turbidity. Moreover, some phytoplankton species, including Microcystis aeruginosa, a freshwater cyanobacteria species, produce harmful toxins, posing risks to both marine aquatic life and human health in this area. Primary nutrient sources in this estuary include watershed freshwater discharges, …

Low Methane Emissions From Nutrient-Rich Former River Channels In A Hydrologically Managed Dryland Delta, Siena Stassi, Jude A. Benavides, Chu-Lin Cheng, Marcelo Gomes Da Silva, Nicholas E. Ray, Nathan O. Barros, Rafael M. Almeida

School of Earth, Environmental, & Marine Sciences Faculty Publications

Anthropogenic stresses in dryland regions can alter hydrology, salinization, and biogeochemical processes. How these pressures influence methane (CH4) cycling and emissions from aquatic ecosystems remains poorly understood. Over the course of 1 year, we measured diffusive and ebullitive CH4 emissions from resacas—shallow, hydrologically managed distributary channels prevalent throughout the Lower Rio Grande Valley along the US–Mexico border. Despite warm, nutrient-rich waters, CH4 emissions from resacas were low, and more comparable to rates observed in estuarine systems. CH4 fluxes were positively related to temperature but negatively related to salinity and rainfall. Elevated salinity, driven by evaporation and source water from the …

The Quality Assurance Machine – A Software Quality Assurance Architecture For Ml-Enabled Systems, Shane E. Downing

All-Inclusive List of Electronic Theses and Dissertations

This dissertation evaluates whether a reusable assurance architecture, the Quality Assurance Machine (QAM), can provide effective product and process quality assurance for ML-enabled software platforms. The QAM is a system-level SQA architecture that turns plans and policies into versioned configurations, executes them in controlled environments, and produces preserved run evidence that supports traceability, auditability, and controlled change. The study follows Design Science Research and evaluates the instantiated artifact using eight assurance requirements (AR1–AR8) synthesized from standards-based guidance, including IEEE 730 and ISO/IEC/IEEE 15026. A four-year longitudinal evaluation combines two methods. First, operational evidence from routine regression and release-validation runs, defect …

Analysis Of Volumetric Reconstruction Methods In Archaeology, Cade O'Fallon

All-Inclusive List of Electronic Theses and Dissertations

The use of Structure from Motion (SfM) photogrammetry in archaeological projects is entering a period of transition; a method of producing 3D data that is traditionally embraced as tool for documentation is being explored for its analytic potential. One such way SfM photogrammetry can be used analytically is through the calculation of volumes using photogrammetric data. The tools exist for archaeologists to be able to create and quantify volumetric models; however, the discourse on these methods is still so new there is no consensus on the best method for conducting volumetric work. Different methods of creating and isolating volumetric space …

Microplastic Composition And Exposure Pathways In Freshwater Systems: A Multi-Matrix Comparison Between Streams And Retention Ponds In Hamilton County, In, Mya Whaley

All-Inclusive List of Electronic Theses and Dissertations

Since the early 1900s, the use of plastics has increased exponentially. The breakdown of plastic materials has exposed all ecosystems to microplastics (small particles of plastic < 5 mm), which are now ubiquitous in natural settings. Microplastics have a relatively large surface area, and many chemicals adsorb to these small particles, impacting their fate and transport of pollution. While microplastics have been found in virtually every setting on Earth, research is necessary to understand their impact on freshwater ecosystems and biogeochemical cycles. The focus of this research was to investigate microplastic compositions in crayfish, sediments, and water samples collected in streams and retention ponds. We analyzed trends in microplastic counts across different land-cover classes, waterbody types, and species, and compared concentrations among matrices to improve our understanding of microplastic exposure pathways in freshwater ecosystems. As a secondary objective, lead and arsenic concentrations were analyzed from sediment samples to investigate metal burdens and compare them to microplastic concentrations. Our findings revealed that, in general, urban streams had the highest microplastic concentrations within land cover categories, and retention ponds had higher microplastic counts than streams. Crayfish, such as F. rusticus had higher microplastic concentrations than F. virilis and F. propinquus, although these results may be influenced by study design. The most common microplastic type was fibers, and most common color was transparent. There was no significant correlation between microplastic and metal concentrations in sediments. We found crayfish to be adequate bioindicators of microplastic pollution, because they were less impacted by environmental variability than sediment and water samples. This research presents the first evidence of microplastics in our study species, F. propinquus, F. rusticus, and F. virlis. This work presents new evidence of microplastic accumulation in crayfish, sediments, and water from freshwater systems in the Midwestern United States. Microplastics were found in every sample type—crayfish, water, and sediment—across urban, rural, and forested landscapes, highlighting their pervasive distribution. The findings provide foundational data for assessing emerging ecological and human-health risks and emphasize the importance of sustained monitoring and research on MP contamination in freshwater environments.

Market Reactions To Deceptive Language In Fake News: Implications From Language Expectancy Theory And Transfer Learning, Ka Chung Ng, Ping Fan Ke, Ping Fan, Mike So, Tam, Kar Yan

Research Collection School Of Computing and Information Systems

The advent of generative artificial intelligence (AI) has heightened the proliferation of fake news. A key challenge is the limited real-world data to investigate the societal impact of fake news produced by generative AI. In this paper, we examine stock market reactions to financial news articles that exhibit stylometric similarity to human-crafted and AI-crafted fake financial news. Grounded in language expectancy theory, we employ a style-based transfer learning model, pre-trained to recognizing deceptive language employed in various types of fake news intricacies. We then apply this model to a comprehensive dataset of financial news, assigning a “veracity style score” to …

Enhancing Traffic Safety Through Ai-Driven, Privacy-Preserving, And Secure Impaired Driving Detection Systems, Razan Alsulieman

Dissertations

Drunk driving remains a major threat to road safety worldwide, contributing significantly to traffic injuries and fatalities each year. Traditional detection approaches are largely reactive and vehicle-centric, relying on in-vehicle sensors, breathalyzers, or post-incident enforcement. These methods often depend on driver cooperation, intrusive hardware installations, or limited monitoring environments, restricting their scalability and effectiveness in large transportation systems. At the same time, modern cities increasingly deploy roadside cameras, surveillance networks, and drone- based monitoring systems, creating new opportunities for proactive intoxication detection at the infrastructure level. However, leveraging such external monitoring introduces challenges related to secure data collection, reliable AI-based …

Mease: Multi-Agent Episodic Action Sequence Explanation, Phyo Wai Khaing, Minghong Geng, Shubham Pateria, Budhitama Subagdja, Ah-Hwee Tan

Research Collection School Of Computing and Information Systems

Multi-agent reinforcement learning (MARL) achieves remarkable performance in complex coordination tasks, yet interpreting the emergent behaviors of trained agents remains a fundamental challenge. Most current explainability methods focus on individual agent decisions, overlooking the critical interplay of joint strategiesand temporal coordination patterns that define successful multi-agent policies. We present MEASE (Multi-agent Episodic Action Sequence Explanation), a novel explainable MARL (XMARL) framework that explains trained MARL policies as human-interpretable emergent cooperative joint behaviors. MEASE employs a cognition-inspired episodic memory model to learn spatio-temporal multi-agent interaction patterns, coupled with abstraction algorithms that identify significant cooperative agent behaviors. We evaluate MEASE on diverse …

Stability Of Paraffin-Insulated Nitrogen-Doped Carbon Fiber Ultramicroelectrodes, Maxwell H. Kwakwah

Electronic Theses and Dissertations

Nitrogen-doped carbon materials can facilitate the reduction oxygen-oxygen bonds, a key step in electroreduction of both O₂ (an important reaction in energy applications) and H₂O₂ (an important reaction in sensing and biosensing). Our lab previously demonstrated that thermal decomposition of urea in the presence of carbon fiber (CF) results in an N-doped carbon material that exhibits good electrocatalytic properties for H₂O₂ reduction. However, previous methods of insulating single N-doped CF via laser heating of borosilicate capillaries and epoxy to prepare ultramicroelectrodes (UMEs) seemed to adversely affect surface nitrogen contents and electrocatalytic properties. More recent work that utilized paraffin to …

Machine-Learning Landslide Susceptibility And Runout Modeling In The Nolichucky River Gorge After Hurricane Helene, Grace Braver

Electronic Theses and Dissertations

Extreme rainfall from Hurricane Helene (September 2024) triggered widespread landslides across the southern Appalachian region, highlighting the need for rapid landslide susceptibility assessments that capture both landslide initiation and downstream runout. Traditional susceptibility models often focus solely on initiation zones, limiting their ability to identify which slopes will generate destructive landslides or where material will travel. This study addresses that gap by (1) integrating Geographic Information System (GIS)-based machine learning susceptibility modeling using ArcGIS Pro: Maximum Entropy (MaxEnt) and Random Forest-Based and Boosted Classification and Regression (FBBC) and (2) the U.S. Geological Survey (USGS) Grfin (Growth, Flow, and Inundation) runout …

Integrated Geophysical, Geotechnical, And Mineralogical Characterization Of Expansive Weches Formation Soils In East Texas: Implications For Infrastructure Design, Samuel Annobil

Electronic Theses and Dissertations

This study investigates the geotechnical, geophysical, and mineralogical characteristics of soils derived from the Weches Formation in Nacogdoches County, East Texas, with emphasis on the relationships among electrical resistivity, moisture condition, compaction state, and mineralogical composition. The Weches Formation is well known for producing clay-rich, moisture-sensitive soils that pose challenges to subgrade performance because of their plasticity and swelling behavior. Representative soil samples were collected and subjected to laboratory testing, including particle size analysis, Atterberg limits, Standard Proctor compaction, free swell tests, electrical resistivity measurements, and X-ray diffraction (XRD) mineralogical analysis. The soils are predominantly fine-grained and clay-rich, with USCS …