Digital Commons Network^™

Towards Robust Deep Learning: Label Noise Modeling, Algorithms, And Applications, Diego Fernando Linares Gonzalez

Graduate Thesis and Dissertation post-2024

This thesis investigates robust and hybrid deep learning with emphasis on improving accuracy and generalization in challenging real-world settings, including learning under noisy supervision and applications to complex scientific data.

In the first part, we address instance-dependent label noise by using a tensor factorization–based methodology that models the noise transition process as an instance-dependent decomposition. We also utilize low-rank block-term structures to capture class-specific and instance-specific correlations. Moreover, we include a robust loss function and rank regularization to achieve stable optimization during model training. We validated this with experiments on benchmark datasets, like Fashion-MNIST, CIFAR-10, and CIFAR-10N, demonstrating higher accuracy …

A Framework For Machine Learning Surrogate Modeling In Physics-Based Reacting Flow Simulations, Jose O. Bobren-Diaz

Graduate Thesis and Dissertation post-2024

The design of next-generation combustion systems relies on computer simulations that capture the multidimensional and multiscale physics of turbulent reacting flows. These simulations are computationally intensive, requiring trade-offs to produce results within allowable time and cost constraints. These trade-offs often consist of general assumptions and modeling simplifications that approximate complex physical phenomena. Consequently, these models can limit both the accuracy and the regime of applicability to real systems since these approximations are often heuristic and inconsistent with the governing physics. Moreover, in today’s rapid design-to-prototype development phase, making timely and informed design decisions is important. This requires fast and reliable …

Advancing Heterogeneous Collaborative Perception For Autonomous Vehicles, Babak Ebrahimi Soorchaei

Graduate Thesis and Dissertation post-2024

Recent advances in computational and communication technologies have enabled deep neural networks (DNNs) and high‐speed vehicular wireless links, paving the way for cooperative perception and cognition in autonomous systems. By exchanging complementary observations, these methods overcome inherent sensor limitations, such as occlusions and restricted range, and thereby improve the detection of partially hidden targets. However, real‐world deployments face challenges from noisy measurements (e.g., GPS inaccuracies) and heterogeneous perception stacks: vehicles often rely heavily on visual detectors, yet there is no standardization ensuring uniform object‐detector architectures across a fleet.

In this dissertation, we first explore the design of a lightweight detector …

Communication-Aware Trajectory Prediction And Tracking For Connected And Autonomous Vehicles, Arash Raftari

Graduate Thesis and Dissertation post-2024

This dissertation presents a unified framework for communication-aware trajectory prediction and vehicle state tracking in connected and autonomous vehicle (CAV) systems. Unlike most deep learning–based forecasting models that assume full observability of surrounding agents, the proposed approach explicitly accounts for limited communication range, packet loss, and bandwidth constraints that characterize real vehicular networks.

A model-based, error-driven information-passing policy is introduced to extend situational awareness beyond one hop. Rather than relaying messages based on distance or freshness, each vehicle evaluates the utility of received content in reducing system-wide position tracking error (PTE). Using compact Gaussian Mixture Model (GMM) representations of motion …

A Multivariable Feedback Approach To Individual Pitch Control For Fatigue Load Reduction In Wind Turbines, Kazi Ishtiak Mohsin Mr.

Graduate Thesis and Dissertation post-2024

This dissertation investigates advanced control strategies to improve performance and reduce fatigue loads in both land-based and floating offshore wind turbines. The proposed approaches target fatigue load mitigation at critical component frequencies while ensuring rotor speed regulation and efficient power generation. The control strategies are implemented and validated using both our in-house modeling and simulation platform, CRAFTS (Control Oriented Reconfigurable Acausal Floating Turbine Simulator), and the industry-standard OpenFAST model.

For land-based turbines, torque actuation and collective blade pitch control are incorporated into a nonlinear controller for rotor speed regulation and power generation. To address fatigue load reduction, individual pitch control …

Attention-Based 3d-Convolutional Neural Network Model For Mechanical Property Predictions Using, Daniel E. Traczyk

Graduate Thesis and Dissertation post-2024

Additive manufacturing (AM), while commonly used for rapid prototyping and creating components with complex geometries, has not been widely adopted for critical applications across the aerospace, automotive, defense, energy, and medical industries. This is, in part, due to the challenges of controlling flaws and uncertainty in the mechanical behavior of additively manufactured components. In recent years, there has been an increase in research aimed at predicting the final mechanical properties of additively manufactured components during the printing process. To address these issues, a 3D-CNN model was trained using low-cost in situ visible-light camera data, anomaly classifications, and the chosen process …

Flywheel Technology: The Reliability Of Functional And Systemic Perturbations In A Crossover Design, Trevor J. Dufner

Graduate Thesis and Dissertation post-2024

The purpose of this dissertation series was to evaluate the reliability, validity, and applicability of the Desmotec D.EVO isoinertial flywheel device (D.EVO) for both isometric and dynamic squat assessments. Study one sought to establish the validity of the D.EVO for assessing squat maximal voluntary isometric contraction (FMVIC) versus a traditional squat maximal voluntary isometric contraction (TMVIC), determine the test-retest reliability of force and power outputs during FMVIC and flywheel squat protocols (FSPs), and explore relationships between these outputs and broad jump (BJ) and vertical jump (VJ) performance. We also identified the number of familiarization sessions needed for reliable data collection. …

Manipulated Workplace Hazing And Attributions, Cole B. Treadaway

Graduate Thesis and Dissertation post-2024

Workplace hazing has scarcely been researched, and when it has, the results have not always aligned. This study seeks to remedy the discrepant findings by examining the importance of hazing severity, as prior operationalizations of hazing severity utilizing the most frequently used and validated measure of workplace hazing do not adequately measure severity, as well as by introducing attribution theory to offer a potential explanation as to why some people respond favorably and others unfavorably to being hazed. This study utilized an experimental 2x2 factorial design with an added control group to examine whether the severity of the hazing and …

Better Models For Mosquito Control: Impacts On Nontarget Insects And Joint Species Distributions Of Mosquitoes, Jacob D. Hart

Graduate Thesis and Dissertation post-2024

Since its inception, mosquito control has been vital for limiting the spread of vector-borne pathogens. Modern control efforts balance the protection of public health and the prevention of environmental harm. We addressed this from two directions: the study of mosquito adulticide effects on nontarget insects, and fine scale distribution modeling of mosquitoes to better target control efforts. We first investigated the effects of ultra-low volume (ULV) pyrethroid applications on nontarget insects in Central Florida and found no statistically clear evidence of reduction in abundance or diversity. We reported a slight and statistically unclear reduction of night-flying Lepidoptera, and effective reduction …

Examining The Impact Of The Menstrual Cycle On Neutrophil Function: Immune Response, Il-8 Sensitivity, And Exercise Perception In Resistance Trained Eumenorrheic Women, Jessica M. Moon

Graduate Thesis and Dissertation post-2024

The purpose of this dissertation series was to examine the effect of the menstrual cycle on neutrophil chemotaxis and adhesion profiles, as well as general immunological, functional, and perceptual measures of recovery following resistance exercise in trained women. Study one explored the impact of the early follicular (EFP), late follicular (LFP), and mid-luteal (MLP) phases on neutrophil recruitment and adhesion receptor expression, markers of muscle damage and inflammation, and functional recovery following resistance exercise. Results indicated that menstrual phase did not have a significant impact on immunological recovery or functional recovery. However, main effects of menstrual phase were revealed in …

Skeletal Embodiment Of Structural Violence: Testing Application Of The Weathering Hypothesis In Analyses Of Identified Sex, Jasper K. Shaw

Graduate Thesis and Dissertation post-2024

Structural violence refers to the experiences of inequality faced by marginalized sex, gender, race, and socioeconomic groups. Unequal access to healthcare, housing, and education can cause poor health that may affect how a decedent is analyzed by forensic anthropologists. This research hypothesizes that, due to gender- and sex-based marginalization, females will display greater rates of stress and correlated accelerated aging (i.e., weathering), than males. Computed tomography scans from 100 males and 99 females aged 20-69 years old and white and middle class, were sampled from the New Mexico Decedent Image Database (NMDID). Stress was assessed via cranial porosities, antemortem tooth …

The Effects Of Resistance Training Volume On Muscle Quality Among Older Adults, Kevan Knowles

Graduate Thesis and Dissertation post-2024

BACKGROUND: Age-related declines in muscle strength are more closely linked to deteriorations in muscle quality than muscle mass. Echo intensity (EI) measured by B-mode ultrasonography is a common non-invasive marker of intramuscular adiposity and fibrous tissue. Although high-intensity resistance training is effective, the added value of higher training volumes on muscle quality and strength in older adults remains unclear. PURPOSE: This study aimed to compare the effects of different resistance training volumes on various measures of muscle quality, size, and strength among older adults. METHODS: Twenty-five older adults were randomized to either moderate volume (2 sets per exercise) or high …

Change Point Monitoring In High-Dimensional Factor Models, Mahdi Mirhosseini

Graduate Thesis and Dissertation post-2024

This dissertation presents the Factor-Augmented Detection (FAD) algorithm, a robust methodological framework for online change point detection in high-dimensional data streams, grounded in classical factor model analysis. To tackle the challenges of monitoring large-scale datasets, the FAD algorithm decomposes data into a parsimonious set of common factors and sparsely changing idiosyncratic residuals, facilitating independent, targeted analysis of low-dimensional factor dynamics and sparse high- dimensional residual evolution.

Following decomposition, structural changes in the factor components are identified using established multivariate sequential monitoring techniques, specifically the Multivariate Cumulative Sum (mCUSUM) and Multivariate Exponentially Weighted Moving Average (mEWMA) procedures. Simultaneously, sparse alterations in …

Effects Of Cerium Concentration And Scattering Centers On The Performance Of Gagg Transparent Ceramic Scintillators, William M. Bowman

Graduate Thesis and Dissertation post-2024

Cerium-doped Gadolinium Aluminum Gallium Garnet (Ce:GAGG) is a rugged scintillator material renowned for its high light output, fast luminescence decay, good stopping power, and energy resolution, making it an ideal candidate for gamma radiation detection. Its optically isotropic crystalline structure enables the fabrication of transparent optical ceramics, effectively addressing certain limitations associated with single-crystal growth. However, the production of highly transparent ceramics often involves costly processing methods with low overall yield. Reactive sintering offers a more economical alternative, but as with other processing techniques, the elimination of scattering centers—such as porosity and secondary phases—remains a significant challenge. These defects hinder …

Holographic Optical Elements For Complex Beam Transformations, Shaghayegh Yaraghi

Graduate Thesis and Dissertation post-2024

In this thesis, we explore the design, fabrication, and application of volume holographic phase elements in photo-thermo-refractive (PTR) glass for two key purposes: expanding the eye-box in augmented reality (AR) displays and controlling optical angular momentum (OAM) modes in laser cavities.

Encoding complex phase structures in PTR glass enables efficient beam shaping, multiplexing, and wavefront engineering, offering strong potential in AR displays, optical communications, and structured light. Holographic optical elements (HOEs) recorded in PTR glass exhibit high diffraction efficiency (over 99%), thermal stability, and chemical durability. Combined with low absorption and scattering in the visible and near-IR, these features make …

Three-Section Multi Quantum Well Mode-Locked Laser For Optical Comb Generation And Space Applicationsical Comb Generation And Space Applications, Di Huang

Graduate Thesis and Dissertation post-2024

This dissertation presents the design, fabrication, and characterization of three-section AlGaInAs/InP multi-quantum-well mode-locked laser diodes (MLLDs) operating near 1.55 µm for optical frequency-comb generation and control in space communication systems. The devices integrate gain, saturable-absorber, and modulation sections, enabling both passive and hybrid mode-locking with precise tunability through independent bias control. Under hybrid mode-locking, the MLLD achieved 1.32 ps pulses with a 5.87 nm optical bandwidth and a 35 dB improvement in RF side-mode suppression compared with passive operation, demonstrating enhanced timing stability and reduced phase noise. Fixed-point analysis established a framework for evaluating dispersion and frequency drift, while the …

Wavefront Sensing Using Nonlocal Metasurfaces, Arturo Martin Jimenez

Graduate Thesis and Dissertation post-2024

Many laser applications require wavefront control to compensate for aberrations caused by system misalignment, component fabrication errors, and propagation through inhomogeneous atmospheres. Aberrations caused by atmospheric turbulence remain particularly challenging to measure due to the temporally fluctuating wavefront distortions and large transverse phase variations in the beam. Conventional wavefront sensors, like the Schack-Hartmann (SHWS) are effective at low turbulence levels, but require relatively bulky optical elements. This thesis investigates the use of metasurface optical elements for use in wavefront sensors both in the low-turbulence limit and in deep turbulence conditions. Such metasurface-based wavefront sensors can achieve the functionality of complex …

Metasurface-Refractive Hybrid Lens Design, Ko-Han Shih

Graduate Thesis and Dissertation post-2024

Refractive optics are widely used in imaging systems while optical aberrations can limit their imaging performance and the typical solution to correct is cascading additional refractive optics with varying materials and shapes. Still, this scheme can result in bulky and costly lenses. Metasurfaces (MSs), with their compactness and ability to locally manipulate wavefronts, offer additional degrees of freedom in aberration correction. Integrating MSs with refractive optics creates MS-refractive hybrid lenses, enabling advanced optical performance while maintaining a compact design. Various methods have been proposed for designing aberration-correcting MSs in hybrid lenses, which often rely on predefined target phase profiles or …

Advancements In Slow-Wave Thin Film Lithium Niobate For Opto-Electronic Communications, Joseph W. Haefner

Graduate Thesis and Dissertation post-2024

The views expressed in this dissertation are those of the author and do not reflect the official policy or position of the United States Air Force, Department of Defense, or the U.S. Government.

The explosive growth of human connectedness over the previous decades has, and will continue to strain the communication infrastructure under-girding modern economies, and society. The continuous expansion of data-centers, advanced artificial intelligence models, and remote data storage will require novel high bandwidth solutions for both long haul and local inter-connects. Current solutions seek to expand link bandwidth through the use of complex high baud rate constellations, in …

Utterance Fluency In Integrated Speaking Tasks: Examining The Role Of Task Type And Topic, Elif Saribas

Graduate Thesis and Dissertation post-2024

Fluency is a fundamental component of second language (L2) proficiency and plays a central role in both communicative competence and language assessment (Tavakoli & Wright, 2020). This study examined utterance fluency, specifically speed, breakdown, and repair fluency, within integrated speaking tasks, widely employed in high-stakes language assessments such as the Test of English as a Foreign Language Internet-based Test (TOEFL iBT). Although previous research has established strong associations between fluency and both communicative adequacy and overall proficiency ratings (De Jong et al., 2015; Révész et al., 2016), the interaction between task structure and topic remains insufficiently understood due to confounding …

Expanding The World Of Themed Entertainment Through Choice-Based Narratives: How Video Game Design Principles Inspire Immersion In Attractions, Layla R. Mccarthy

Graduate Thesis and Dissertation post-2024

Within most traditional theme park attractions, guest interaction does not directly impact the unfolding of the narrative. While there are many examples of encouraged interaction, guests still lack agency. Within this research, the goal is to develop a hypothetical theme park attraction (titled The Library of Worlds) where guests will be met with various choices that influence the trajectory of the story. Drawing inspiration from non-linear storytelling and interactive gaming, this study explores integration of choice-based narratives into themed entertainment, questioning the ways in which agency can enhance immersion and connection within attractions – as well as what elements …

Evaluating Robustness Of Methods For Comparing Two Means Using Paired Data With Few Incomplete Pairs, Thien Pham

Graduate Thesis and Dissertation post-2024

Practitioners often resort to the paired t test to test equality of two means when data is complete. When small amounts of data are missing, is this method still as robust? Are there alternative methods which can perform the job better on the missing data? In this thesis, we measured the robustness of the paired t-test, two sample t-test, corrected z-test, modified corrected z-test, weighted t-test, pooled t-test, optimal pooled t-test, unequal variance test, correlation focused t-test, multiple imputation method, mixed model method, modified maximum likelihood estimate method, and Uddin’s modified maximum likelihood estimate method. This is simulated in an …

Under Control, Eric Tegethoff

Graduate Thesis and Dissertation post-2024

Under Control is a collection of six short stories focused on masculinity. The stories explore family, climate change, nationalism, and the nature of work. “A Survey of Man” and “Snake Pit” dive into the corrosive effects of capitalism on men in society. In “The Art of War,” the narrator finds the cost of pursuing healing outside of society’s prescribed norms. Society’s norms also act on the characters in “My Country, My Cybertruck” and “Bad Reporting.” The collection ends with “Virtual Idaho,” a short story that scrutinizes the masculine barriers to properly grieving for a loved one.

Of Fate And Fear: An Interactive Narrative Approach To Experiential Horror, David A. Brescia

Graduate Thesis and Dissertation post-2024

This thesis examines the nature of guest interactivity in themed entertainment, considering best practices when approaching the creation of an interactive theme park offering with the belief that narrative interactivity and guest agency increases the potential for meaningful, memorable, and engaging entertainment experiences. Taking both established literature and existing offerings into consideration, this thesis presents an original creative narrative design concept for a Halloween haunt event overlay, titled Fear the Unknown, that enhances the guest experience with an emphasis on interactivity and agency. This creative work is presented specifically from the point of view of the show writing discipline, consisting …

Millennials And Generation Z Customers' Service Recovery Evaluation From A Self-Esteem Perspective, Megan Mcdade

Graduate Thesis and Dissertation post-2024

While a significant amount of research has been conducted to assess the effectiveness of service recovery based on various factors, there is limited research on emerging customer segments, specifically Generation Z and Millennials. Justice theory has been predominant in service recovery literature, however, recent studies suggest that self-esteem can provide valuable insight into understanding customers’ evaluations of service recovery. This research focuses on how Generation Z and Millennial customers evaluate service recovery efforts in the context of restaurants. The main objective of this study is to examine how these customers perceive economic (compensation) and psychological (apology) service recovery strategies, considering …

Understanding The Mechanism In Shaping Guest Acceptance Of Service Robots In Caribbean Hotels: The Moderating Role Of Cultural Familiarity In The Cac Model, Jamilia A. Lawrence

Graduate Thesis and Dissertation post-2024

Service robots continue to become increasingly popular in the hospitality and tourism industry. Understanding the cognitive and emotional processes is important in explaining the sequential processes of guests accepting new technologies. Especially in regions like the Caribbean, where human-to-human interaction and personalized services are highly valued.

This study investigates how cognitive evaluations of service robots, including utility, coolness, autonomy, courtesy, and interactivity, influence guests' emotional responses and their acceptance of the technology in Caribbean hotels. Additionally, the study explores the moderating role of cultural familiarity in the relationship between emotion and acceptance.

Using the cognitive-affective-conative (CAC) model, 415 responses were …

Comparing Inter-Rater Reliability For Two Coding Methodologies In Meta-Analysis Data Extraction, Christine D. Destefano

Graduate Thesis and Dissertation post-2024

Meta-analyses are critical tools for evidence-based policymaking, but their validity depends heavily on accurate data extraction from primary studies. While previous research has focused on procedural aspects of coding and reliability of effect size extraction, there remains a significant gap in understanding how to improve inter-rater reliability (IRR) for moderator variables that provide crucial context for meta-analytic findings. This study compared two methodological approaches for coding moderator variables: traditional low-inference coding and a novel drill-down questioning technique adapted from database development principles. Using a comparative analysis design, two independent coders extracted data from 18 primary studies included in a published …

Characterizing Conditioned Pain Modulation And Quadriceps Motor Unit Recruitment Among Older Adults With Chronic Knee Pain, Emily J. Parsowith

Graduate Thesis and Dissertation post-2024

BACKGROUND: Over 25% of U.S. older adults experience chronic knee pain, which worsens with inactivity, creating a cycle of pain, disability, and sedentary behavior. However, a limited understanding of its underlying mechanisms hinders the development of effective treatments. PURPOSE: This study integrated assessments of Conditioned Pain Modulation (CPM) and motor unit control to elucidate the role of pain sensitization and neuromuscular impairments specific to chronic knee pain. METHODS: Seventy-five participants were divided into three groups: young adults (n=25), older adults without pain (n=30), and older adults with chronic pain (n=20). CPM efficiency was evaluated using heat and pressure test stimuli …

Understanding Pre-Service Teachers’ Engagement, Application And Perception Of Instructor Feedback In An Online Teacher Education Course: An Action Research Study, Iram Khan

Graduate Thesis and Dissertation post-2024

Teachers should prioritize continuous professional development through shared experiences and reflective teaching to remain competitive in the global education system and become better teachers to assist language learners in achieving their goals (Vadivel et al., 2021). Professional development courses are crucial for enhancing teaching skills; however, there is a concern that such initiatives seem ineffective in promoting change in teaching practices (Darling-Hammond et al., 2017; Gore et al., 2017). Implementing effective instructional approaches in teacher education courses, such as feedback and reflection, is crucial for aspiring teachers to enhance their knowledge and professional growth (Darling-Hammond et al., 2017; Pharis et …

Florida Title I School Principals' Characteristics And Practices In Professional Learning As Related To Student Reading Achievement, Sarah K. Hall

Graduate Thesis and Dissertation post-2024

This study explored the relation between Florida’s Title I elementary school principals’ characteristics and how they design professional learning with student reading achievement. This study used Descriptive Statistics, Spearman’s Correlation, Independent Samples T-tests, Pearson’s Correlation, and ANCOVA to explore the relation between 259 of Florida’s Title I public elementary school principals’ characteristics (years of experience, number of state reading certifications, degrees, and level of autonomy) and how they design professional learning with student reading achievement. The research design was quantitative and correlational. Data was collected from 58 districts through a piloted survey using parametric sampling. A statistically significant difference was …