Physical Sciences and Mathematics Commons^™

Advancing Multi-Agent Robotics Simulations Through Heterogeneous Reinforcement Learning In Isaaclab, Jacob R. Haight

All Graduate Theses and Dissertations, Fall 2023 to Present

Robots increasingly operate in collaborative teams across domains such as search-and- rescue, warehouse automation, and autonomous driving—scenarios that demand advanced coordination strategies enabled by multi-agent reinforcement learning (MARL). However, existing simulation frameworks often struggle to balance realism, speed, and scalability, especially when supporting diverse, heterogeneous robot teams. This research extends Isaac Lab, a high-performance robotics simulator, by integrating heterogeneous-agent reinforcement learning (HARL) capabilities. The result is a flexible and GPU-accelerated platform for training both homogeneous and heterogeneous robot teams in complex, physics-based environments. These enhancements significantly narrow the gap between simulation and real-world deployment for multi-robot systems.

Development Of Saddlepoint Methodologies For Sparse Sample Multiple Parameter Generalized Linear Models And Correlated Data Scenarios, Christopher Johnson

All Graduate Theses and Dissertations, Fall 2023 to Present

Parameter estimation using maximum likelihood techniques may be biased when sample sizes are small, event rates are low, or otherwise sparse counts exist in a parametric model. This in turn may lead researchers to draw invalid statistical conclusions when conventional methods are utilized. The saddlepoint approximation has potential to lessen the degree of bias in sparse data conditions through its use of moments beyond the mean and variance, which allows for more accurate approximations using a smaller number of observations. We propose two novel saddlepoint methods for use in practical analysis scenarios, as an alternative to maximum likelihood estimation. First, …

Determining The Structure And Function Of Type Iv-A Anti-Crisprs, Olivine Redman

All Graduate Theses and Dissertations, Fall 2023 to Present

Bacteria face a constant existential threat in the form of infection by viruses along with other forms of mobile genetic elements, such as bacteriophage and transposable elements. To survive, bacteria and other prokaryotes have evolved various immune systems to evade these would-be invaders. One such immune system is the CRISPR-Cas system, an adaptive immune system able to record the genetic signature of invading viruses in order to recognize and destroy them should they be encountered again in the future. In this thesis I present data that sheds light on the mechanism of one particular subtype of CRISPR-Cas systems: the type …

Conservation Laws For Asymptotically Perfect Fluid Spacetimes, Lyle Arnett Jr.

All Graduate Theses and Dissertations, Fall 2023 to Present

In physics, conservation laws, like those for energy or momentum, are powerful tools that help us understand how systems evolve and interact. In general relativity, where spacetime itself is curved by matter and energy, defining such conservation laws becomes especially subtle and complex. Traditional methods rely on idealized conditions like empty space or special symmetries, which limit their usefulness in more realistic, dynamic settings such as an expanding universe.

This dissertation explores a modern mathematical framework, developed by Iyer and Wald, that allows conservation laws to be derived directly from the equations governing spacetime. Using this approach, I examine not …

Zero Waste Initiatives Across Three Institutions: Local Government, An Academic Institution, And A Retail Business, Soren G. Gray

Graduate Student Portfolios, Professional Papers, and Capstone Projects

No abstract provided.

Ml Playground: Data Modification/Preprocessing And Model Simulation Tool, Marco D. Cerrato

Electronic Theses, Projects, and Dissertations

There is a heavy reliance on programming when it comes to learning machine learning (ML). This often creates barriers for students and newcomers unfamiliar with coding. While the lessons you learn in the classroom provide essential foundational understanding, some technical or practical aspects of ML—such as data preprocessing, feature engineering, and model tuning—are best learned through hands-on interaction. ML Playground was developed to act as a proof-of-concept application to address this gap by offering a browser-based, graphical user interface that lets users engage with core ML workflows without writing code. Designed with educational accessibility in mind, the application allows users …

Sediment And Debris Flows Resulting From The 2020 El Dorado Wildfire, San Bernardino Mountains, California, Andrew Suarez

Electronic Theses, Projects, and Dissertations

The Yucaipa Ridge is a section of the San Bernardino Mountains within the Transverse Ranges of Southern California. During the summer and early fall of 2020, Yucaipa Ridge experienced severe vegetation damage from both the Apple and El Dorado wildfires. Burned slopes exhibit a higher susceptibility to geomorphic change within several years of burning from the introduction of meteoric water onto the slope. Yucaipa, Oak Glen, and other nearby communities are currently at risk of damage by hyperconcentrated flows and debris flows during the seasonal rainy season.

Debris flows and hyperconcentrated flows (a.k.a. mudflows; are grouped and referred to in …

Geochemistry Of The Inyo Volcanic Chain, And Evaluation Of The Portable Niton Xrf Instrument, Dylan Terry

Electronic Theses, Projects, and Dissertations

The Inyo Volcanic Chain (IVC) is a series of rhyolitic lava domes straddling the northwest rim of the Long Valley Caldera (LVC), most recently erupting ~650 years ago, producing the South Deadman, Obsidian, and Glass Creek Domes. For this study, samples were analyzed for geochemistry and petrography at six of the IVC domes. To measure geochemistry, a portable x-ray fluorescence machine (PXRF) was used, in part to test how well it performed on felsic rocks. The PXRF performed poorly with the factory calibration in detecting most elements, though detection of some elements improved with a calibration curve applied. The three …

A Mineralogical And Geochemical Investigation Of The Influence Of Tectonic Setting On Accessory Mineral Assemblages In Serpentinites Along The Western North American Margins, Bryan H.T. Seymour

Electronic Theses, Projects, and Dissertations

Serpentinites are metamorphic rocks typically produced by hydrating mantle peridotites to form assemblages containing one or more serpentine minerals. They occur in various tectonic and geologic settings, such as submarine hydrothermal systems, ophiolite sequences, and the forearc mantle. Serpentinites are associated with highly reduced conditions, as indicated by low oxygen fugacity (ƒO₂) values. Previous investigations suggest that oxygen fugacity varies with tectonic setting due to differences in environmental conditions. To test this hypothesis, we analyzed serpentinites from a wide variety of tectonic settings in western North America: The New Idria forearc diapir (CA), Canyon Mountain Island arc complex (OR), Josephine …

Numalyze: Numerical Analysis Web Application, Dev Kapupara

Electronic Theses, Projects, and Dissertations

Numalyze is an online platform that allows users to run and apply different numerical methods in real time. The application is built using Python and the Flask web framework. It provides an interface where users input mathematical functions and parameters to see the results for root-finding and integration methods, and to also perform reductions on matrices. By using a light-weight web framework and self-coded algorithms which removes dependency on massive external libraries—this application connects theoretical concepts to their practical implementation. It enables students and researchers to visualize the series of steps that each algorithm takes to compute results. Moreover, the …

Understanding The Impact Of Ransomware On Biotechnology, Tswvyim Vang

Electronic Theses, Projects, and Dissertations

Biotechnology encompasses the use of research on both biology and technology to create products that can advance in areas such as healthcare and agriculture. Because of the technology and products that arise from the use of biotechnology, the industry is especially targeted by cyber-attacks. A prominent type of cyber-attack that is utilized by cyber criminals on biotechnology is ransomware. The goal of this project is to determine the impact of ransomware on biotechnology. The research questions posed are: Q1) What are real examples of ransomware attacks that have occurred on biotechnology companies and what patterns can be identified by these …

Constructing Binary Encoding Matrices From Joined Graphs, Joshua Avalos

Electronic Theses, Projects, and Dissertations

Codes and technology are part of our daily lives and allow the modern world to function, and for us to have conveniences in our lives such as smartphones that can be used to privately call people on the other side of the planet, and for secure access to the internet. In this thesis we will explore the construction of binary codes created by vertex-edge incidence matrices of planar graphs. The Hamming (7,4) code was an incredible code that allowed the detection and correction of errors after receiving them through a transmission. We will explore the possibility of the creation of …

Explainable Ai (Xai) For A Machine Learning Heart Disease Prediction Model, Sai Abhishek Sanchula

Electronic Theses, Projects, and Dissertations

Cardiovascular diseases (CVDs) remain the leading cause of mortality worldwide, necessitating the development of accurate and interpretable machine learning (ML) models for early diagnosis and risk assessment (World Health Organization, 2021). While ML algorithms such as logistic regression, decision trees, support vector machines (SVM) (Cortes & Vapnik, 1995), and deep learning models (LeCun et al., 2015) have demonstrated high predictive accuracy, their adoption in clinical practice is hindered by their black-box nature (Rudin, 2019). Explainable AI (XAI) techniques, including SHapley Additive Explanations (SHAP) (Lundberg & Lee, 2017), Local Interpretable Model-agnostic Explanations (LIME) (Ribeiro et al., 2016), and feature importance analysis …

Time Series Deep Learning Approach For The Intermittent Operational Performance Of A Wellhead Water Treatment And Desalination System, Michael G. Clement

Electronic Theses, Projects, and Dissertations

Distributed water treatment and desalination (DWTD) systems are becoming significant for serving disadvantaged communities that are geographically segregated from centralized water distribution networks. However, given the remote nature of the communities, these systems must operate autonomously adapting to intermittent operations due to varying water use patterns and unavailability of continuous manual labor support. Machine Learning models describing and forecasting system performance are critical, allowing for model-based control, performance forecasting, fault detection, and determination of causal relationships among process attributes. Accordingly, graph convolutional neural networks with an attention mechanism (GATConv) were developed to describe the intermittent operational profiles of a wellhead …

On Bed Posture Recognition Using Deep Learning With Pressure Sensors, Farheen Akhter Ms

Electronic Theses, Projects, and Dissertations

In healthcare applications such as disease prevention, sleep quality evaluation, and patient monitoring, bed posture recognition is essential. Using pressure sensor arrays placed on top of or embedded in mattresses, this study investigates the application of deep learning models for non-invasive posture classification. Although they have been widely employed, traditional machine learning approaches like support vector machines (SVM) and k-nearest neighbors (KNN) sometimes struggle with feature extraction and real-time performance necessitating considerable processing resources. I implemented a model using conventional approaches to get over these restrictions, then fine-tuned it using the following deep learning architectures for bed posture recognition: ResNet-50, …

Degradation Of Martian Glacier-Like Forms In Relation To The Observed Evolution Of Emmons Glacier On Mount Rainier, Wa, Jose E. Jimenez

Electronic Theses, Projects, and Dissertations

This study establishes parallels between the observed degradational evolution of debris-covered glaciers on Mount Rainier, WA and select glacier-like forms (GLFs) by studying the time-varying morphologies of the debris cover. Mount Rainier is home to 28 debris-covered valley glaciers, including Emmons Glacier which has a history of orthoimages taken from 1951 to 2023 and high-resolution Digital Elevation Model (DEM) coverage of 2008, 2021 and 2022. We can observe the degradational evolution of Emmons Glacier through orthorectified black and white imagery collected from an airborne platform and the National Agricultural Imagery Program (NAIP) colored satellite images periodically collected over the last …

Geochemical Analysis For Potential Critical Mineral Resources In Carbon Fly Ash, Korei D. Patterson Teer

Graduate Theses and Dissertations

The growing demand for critical minerals, coupled with the increasing supply chain vulnerabilities, has intensified the need for alternative domestic resources of critical minerals beyond traditional mining. Many of these critical minerals are essential for advanced technologies, energy storage, and national security, yet the United States remains heavily dependent on foreign imports, particularly from China. This study evaluates the economic potential of critical mineral recovery from Carbon Fly Ash (CFA), a byproduct of coal combustion, to determine its viability as a secondary source of critical minerals. A geochemical and mineralogical assessment was conducted on CFA samples from various storage sites …

Exploring Latent Mediation Through Bayesian Regularization Methods Of Lasso, Ridge, Horseshoe, Spike-And-Slab, Ethan Harris

Graduate Theses and Dissertations

Regularization is a powerful tool to combat overfitting and drive sparsity in complex models. Regularization was initially applied in regression modeling but has been increasingly utilized in structural equation modeling where its utility in identifying the essential components has helped improve modeling. As structural equation models have increased in complexity both in the number of indicators but also the number of latent factors, researchers have begun to investigate how applying Bayesian regularization to these systems can further push the limits on modeling complex models with limited sample sizes. One area where research is limited is the application of Bayesian regularizations …

Contrastive Learning Techniques For Fraud Detection, Vinay Madanbhavi Shashidhar

Graduate Theses and Dissertations

Detecting fraud in computing platforms involves identifying malicious user sessions, often using deep learning models, but several challenges hinder effective deployment. Attackers can craft diverse malicious sessions that closely resemble normal ones, complicating the learning of robust decision boundaries. While supervised contrastive learning offers a promising solution through class-specific clustering, its potential remains underexplored. Real-world datasets typically contain few labeled malicious sessions and many normal ones, creating an open-set anomaly detection challenge. Costly expert annotation further limits labeled data, especially for smaller organizations, leading to Positive Unlabeled (PU) learning and noisy label learning issues. Organizations are increasingly turning to LLMs …

Solving Real-World Optimization Problems Using Near-Term Quantum Computing With Applications In Vehicle Routing And Drone Delivery, James Bradley Holliday

Graduate Theses and Dissertations

Quantum computing (QC) stands at the cusp of revolutionizing computation, yet its near-term potential, constrained by Noisy Intermediate-Scale Quantum (NISQ) devices, remains underexplored. This dissertation investigates how hybrid quantum-classical algorithms can address combinatorial optimization challenges in logistics, focusing on vehicle routing and drone delivery—NP-hard problems with exponential solution spaces that defy classical exhaustive methods. Amidst NISQ limitations like limited qubits and high noise, we confront key challenges: encoding complex constraints, e.g., time windows, battery capacity, into quantum models, balancing quantum and classical components for scalability, and accessing scarce quantum resources. By integrating quantum annealing (QA) and the Quantum Approximate Optimization …

Learning Behaviors In Physics-Informed Deep Learning, Alex Glover

Electronic Theses and Dissertations

Physics-informed deep learning is a methodology in artificial intelligence aimed at combating the large training data requirement and the barrier of domain awareness that deep learning architectures commonly face in applications. Stochastic modeling integrated into the predictive models provides that domain knowledge. Variations of the Intelligent Driving Model impact the learning behaviors of the joint-training architecture. This thesis examines the effect of substituting the standard linear Intelligent Driving Model with a modified nonlinear version, as applied to real human driving behavior on the I-80 interstate. The experimentation also critically evaluates the complications that impede the viability of this architecture in …

Kinetic Analysis Of Hydrolysis And Hydrothermal Conversion Of Methyl Cellulose To Dodecane, Joseph Olembo Were

Electronic Theses and Dissertations

Environmental pollution due to emission of greenhouse gases and increasing global energy demand are some of the concerns associated Analysis with the use of fossil fuels. These concerns have triggered the exploration of sustainable and efficient methods for renewable energy production. One such method is the conversion of cellulosic biomass to hydrocarbon fuels using hydrogen bronze metals. Cellulose is the most abundant biomass globally, and thus its utilization can lead to more sustainable energy production. Previous studies have reported the conversion of cellulose to hydrocarbons at high temperatures and hydrogen pressure. This work focuses on the conversion of methyl cellulose …

A Kinetic Analysis On Hydrolysis And Hydrothermal Conversion Of Cellulose To Dodecane, Joel Nyangoka Onyancha

Electronic Theses and Dissertations

Cellulose, the most abundant biomass on Earth, offers a promising pathway for alternative energy through conversion into alkane hydrocarbons. Traditional methods rely on high temperatures and high-pressure hydrogen, often derived from petroleum. In this study, we investigated the conversion of Microcrystalline Cellulose using hydrogen metal oxides (bronzes) under milder conditions. Following ASTM E3417-24 hydrolysis protocols, we employed hydrogen bronzes with toluene for product extraction. This approach enabled conversion at lower temperatures without pressurized hydrogen. Products were analyzed using Gas Chromatography Mass Spectrometer for identification and Gas Chromatography Flame Ionization Detector for quantification. Our findings revealed the formation of a possible …

Analyzing Meteor Shower Radiants, Elise Springman

Celebrating Scholarship and Creativity Day (2018-)

Meteor shower origin points, or radiants, are a widely studied topic in astrophysics as they give a lot of context to the meteor shower itself as well as provide a way study broader space. Specifically, when it comes to annual showers, these radiant points are compared with previous years to create a broader database about the shower and check accuracy of instruments. This project is ongoing with several places tracking these annual showers and comparing data to give a more detailed history of each shower. Here at CSBSJU, meteor showers were analyzed using an AllSky camera, and a few additional …

Impact Of Shear Flow On Marine Biofilms Using Hyperspectral Imaging, Shantanu Mahesh Kore

All Theses

Various microbial communities in marine biofilms cause biofouling on submerged surfaces, posing challenges to marine industries. Despite their ecological and economic importance, biofilms' biomechanical and biochemical responses to hydrodynamic shear stress are insufficiently understood, particularly in dynamic flow conditions. To fill this gap, our study uses an innovative fiber-optic hyperspectral imaging (HSI) system and a supercontinuum laser source to examine marine biofilms' structure, composition, and detachment behavior at different shear stress levels.

To detect spectral and spatial heterogeneity in live biofilms without labeling, we created a custom imaging pipeline that captures reflectance spectra in the 600-850 nm range, targeting microbial …

Application Of Regression Techniques On Designed Economic Data, Naomi O. Edegbe

All Theses

Evaluating stock market data and public companies' performance is an overwhelming task for day traders and brokers in the United States and internationally. As a financial metric of a company's overall valuation, earnings per share is a commonly researched measure of a company's profitability. We investigate relationships between earnings per share, multiple financial measures reported from company income statements, and classifiers such as market capitalization and sector. Multiple linear regression models are developed and assessed for this data. Results conclude that there is a significant difference between sectors and earnings per share recorded for a given company. Individual stock analysis …

Evaluating The Role Of Electrostatic Forces In The Adsorption Of Perfluoroalkyl Substances To Granular Activated Carbon, Thomas A. Mccall

All Theses

In response to new drinking water regulations announced by the EPA in 2024, many municipalities will need to monitor concentrations and implement new treatments to address PFAS, a class of anthropogenic organic contaminants. A leading technology for this application is GAC adsorption due to its familiarity and availability. The adsorption of four PFAS species (PFBA, PFBS, PFOA and PFOS) to Calgon Filtrasorb400 (F400) was evaluated for a range of pH and ionic strength conditions to investigate the role of electrostatic interactions in adsorption capacity. These PFAS are all strong acids, so pH was used as a control for GAC surface …

Novel Methods For Assessing And Prioritizing Road-Stream Crossings For Aquatic Organism Passage, Lesley E. Twiner

All Theses

Anthropogenic barriers such as dams and culverts have led to riverine habitat fragmentation and decreased fish diversity worldwide. Low-head structures such as culverts are far more abundant than large dams and may have a larger cumulative impact on river connectivity. Many efforts to inventory road crossing barriers and prioritize removal and restoration efforts do not consider temporal variation in flow. We used cameras to gain continuous water level monitoring data to quantify the relationship between precipitation events and outlet drop height. In the spring of 2024, we selected 25 culvert sites in upstate South Carolina and set up cameras to …

Algebraic Properties Of Boolean Models, Harrison Fisher

All Theses

Boolean models are n-tuples of polynomial functions in n variables over the finite field of order 2. These models define finite dynamical systems which are used for modeling many different biological systems such as gene regulatory networks. These systems can be defined by updating every function synchronously, or by updating one function at a time asynchronously. In this project, we discuss a method for reverse engineering the model space of all Boolean models which fit a set of partial asynchronous data. This method is a generalization of a known method for synchronous data. In addition, we show that given the …

Divergence-Free Smoothed Particle Hydrodynamics In A Stream Digital Twin, Austin Hartley

All Theses

Digital Twins (DT) are being explored by the South Carolina (SC) water community to simulate how SC streams will flow at various water levels. Currently, a DT called Gilligan simulates these streams utilizing weakly-incompressible Smoothed Particle Hydrodynamics (SPH). This method does not strictly enforce incompressibility, which leads to unrealistic water flows and unwanted visual artifacts that require post-processing effects to hide. To address these problems and simulate more realistic water flows, the Gilligan stream logic is updated and a state-of-the-art SPH method that enforces incompressibility—Divergence-Free SPH (DFSPH)—is implemented within the Gilligan framework. DFSPH is able to make use of two …