Digital Commons Network^™

Parameter-Efficient Multimodal Adaptation: Ocr-Integrated Lora For Textvqa And Captioning, Karthik Ganesh Malini

Master's Theses

Vision-Language Models (VLMs) have emerged as transformative technologies for multimodal AI, yet they face significant hurdles in processing text-rich images required for enterprise applications like document understanding, medical imaging, and industrial inspection. Current VLMs struggle with accurate text extraction and reasoning, often exhibiting high hallucination rates and poor Optical Character Recognition (OCR) token utilization. To address these limitations, this research presents a comprehensive framework for optimizing parameter-efficient Low-Rank Adaptation (LoRA) fine-tuning strategies on state-of-the-art architectures, including LLaVA-1.5 and BLIVA-FlanT5. Our methodology integrates enhanced OCR token utilization, faithful caption generation, and specific hallucination mitigation techniques. We employ a multi-dimensional evaluation protocol …

A Functional Mri Meta-Analysis Encoding In Alzheimer’S Disease, Joanne Ha

Master's Theses

Alzheimer’s disease (AD) is the most common form of dementia and is characterized by episodic memory impairment. Numerous functional magnetic resonance imaging (fMRI) studies have examined brain activity differences between individuals with AD and age-matched controls during memory encoding. However, small sample sizes and methodological variability across studies have limited the ability to draw meaningful conclusions regarding brain activity underlying encoding impairments. We conducted a meta-analysis to identify consistent patterns of hyperactivity and hypoactivity in AD during episodic memory encoding. Using PubMed to identify studies that met our inclusion criteria, 1,002 studies were screened; ultimately, seven studies were included in …

Building A Data-Driven Security Ai Framework Using Machine Learning Models, Christopher Chan Vi

Master's Theses

This research project explores a modern approach to Intrusion Detection System (IDS) anomaly detection by leveraging Artificial Intelligence (AI), Machine Learning (ML), Large Language Models (LLM), and Explainable AI (XAI). The purpose is to assess the effectiveness of these technologies in enhancing the understanding of intrusion events. This research study addresses the challenges of threshold determination and the interpretability of anomaly detection results. The proposed solution involves an LLM-based framework with XAI capabilities, integrated with a Retrieval Augmented Generation (RAG) architecture, to provide clear explanations for detected anomalies, utilizing both custom and pre-trained datasets. The study navigated inherent challenges, including …

Fully Integrated Slippage Detection System For Lower Limb Amputees, Christopher E. Miglio

Master's Theses

Lower limb amputees face significant challenges in maintaining a proper prosthetic fit, as improper fit can lead to slippage at the limb-socket interface, resulting in discomfort, pressure sores, and long-term musculoskeletal complications. To address this issue, a fully integrated slippage detection system was developed to monitor an amputee’s daily activities and slippage occurrences to understand their prosthetic fit over time. The system consists of a prosthetic sock embedded with Interlink 406 flexible piezoresistive force-sensing resistors (FSRs). The sock, worn directly on the residual limb, continuously measures pressure at the limb-socket interface. Sensor data from six FSRs is sampled at approximately …

Self-Efficacy In Juvenile Probation Officers: An Examination Of Training, Eva K. Ott Hill

Master's Theses

The role of juvenile probation officers (JPO) has significantly changed in the past two decades. Rather than a focus on compliance, juvenile probation officers are tasked with using evidence-based practices focused on rehabilitation. Training is a key tool as it equips juvenile probation officers with the skills they need to implement new evidence-based practices and programs. Juvenile probation officers’ confidence in their ability to complete their changing job tasks should increase when they receive training. The current study examines how training impacts perceived self-efficacy among juvenile probation officers in Connecticut. The study seeks to understand what aspects of training impact …

Using Noninvasive Calibrated Cuff Plethysmography To Observe The Effects Of Lifestyle Habits On Arterial Compliance, Joshua M. Whitaker

Master's Theses

Cardiovascular disease remains as a leading cause of death for both sexes and across all racial backgrounds. Continuing to research causes, early predictors, and preventative actions are paramount to ensuring that cardiovascular disease starts to decrease in its activity. Endothelial dysfunction is the earliest known predictor of cardiovascular disease, therefore investigating early detection methods of endothelial dysfunction can improve health outcomes. This study revolves around a non-invasive calibrated cuff device that uses the principles of segmental plethysmography and oscilliometry to measure arterial compliance of baseline and induced hyperemic conditions in human subjects to determine how well arteries vasodilate in response …

Exploring Synthetic Data Generation For Machine Learning And Statistical Analysis, Aditi S. Gajjar

Master's Theses

In the data-driven machine learning and statistical analysis world, real-world data availability is often constrained by privacy concerns, access restrictions, or resource limitations. To address these challenges, synthetic data provides a groundbreaking solution, enabling academics and data scientists to generate artificial datasets that mimic real-world data for model development and testing in a controlled, privacy-preserving environment. The syntheticDatasets package represents an R-based counterpart to Python's widely-used Scikit-learn dataset generators, adapted to meet the needs of statisticians, educators, and data scientists primarily working in the R programming ecosystem.

This thesis explores the design and implementation of the syntheticDatasets package, focusing on …

Performance Enhancement For Rufa: Rapid Urban Forest Assessment, Nicholas Tan

Master's Theses

Urban forests are crucial to the livability and resilience of cities, offering critical ecosystem benefits such as air quality enhancement, temperature regulation, and biodiversity. Managing said urban forests is essential to ensure their sustainability and adaptability to rapidly changing environmental and climate conditions. The Rapid Urban Forest Assessment (RUFA) tool was developed to address the need for a standardized approach to evaluating and comparing urban and community forestry programs. By analyzing and aggregating tree-specific data across California, such as canopy cover, tree counts, and diversity scores, RUFA assigns a comprehensive urban forestry score for each city. This score allows for …

Flower Color Variation And Its Maintenance: Experimental Ecology Of Serpentine-Adapted Leptosiphon Parviflorus Benth. (Polemoniaceae), Magdalene S. Lo

Master's Theses

Environmental heterogeneity can impose fluctuating selection pressures on species across space and time. If variations of a trait provide fitness benefits in certain environmental conditions but not others, these fitness tradeoffs under fluctuating selection could be a means of maintaining intraspecific variation. A common type of intraspecific variation is the production of anthocyanins, pigments responsible for red, pink, and blue colors in plants and that can also provide protection from abiotic stressors, such as water limitation and heavy metals. Leptosiphon parviflorus Benth. (Polemoniaceae, variable linanthus) is an annual wildflower whose individuals produce either pink or white flowers, and both flower …

Using Scenario-Based Design Activities To Explore Sociotechnical Thinking In Engineering Professionals, James M. Whealan

Master's Theses

Sociotechnical thinking is the ability to recognize the interconnectedness of both technical and non-technical dimensions such as environmental, economic, social, cultural, and political factors. Although traditional engineering curriculum often emphasizes technical knowledge and skills, research suggests that engineering work is inherently sociotechnical. While sociotechnical thinking is a critical engineering skill, relatively little is understood about the nature of it or the different aspects comprised within it. To that end, this study addresses the following research question: What sociotechnical factors do engineering professionals consider in a collaborative scenario-based design activity? To answer this question, we conducted a qualitative study of engineering …

Analyzing Urban Soil Moisture Profiles: Implications For Stormwater Management In New Haven, Ct, Tushar Ahmed

Master's Theses

Effective stormwater management is essential for enhancing the resilience of urban environments facing challenges from climate change, aging infrastructure, and increasing impervious surfaces. As cities adopt green stormwater infrastructure such as bioswales to relieve these pressures, understanding the hydrologic behavior within these systems becomes essential. This study investigated soil moisture dynamics in New Haven, Connecticut by analyzing the seven critical points (A, B, C, D, E, F′, and F) that describe pre-storm, during storm and post storm drying behavior within bioswale systems. A total of 77 bioswale sites were monitored during May and June 2025, covering 14 storm events. Soil …

The Impact Of Using The Visual, Auditory/Aural, Reading/Writing, And Kinesthetic (Vark) Model In Ballet Class: A Multifaceted Approach To Teaching Ballet Technique To Children Ages Nine To Twelve Years Old, Angelle Cain Wise

Master's Theses

The purpose of this study was to generate data that would measure the impact of the application of the Visual, Auditory/Aural, Reading/Writing, and Kinesthetic (VARK) Model learning styles on student comprehension and execution of movement in ballet class. The researcher sought to answer these questions:

Q1 Does teaching students’ using their preferred learning style help them understand the movement thoroughly?

Q2 Does teaching students’ using their preferred learning style help them execute the movement more precisely?

There was a total of twelve participants ages nine to twelve years old and two observers, whom were colleagues of the researcher. A set …

So You Think You Can Teach Dance?: Evaluating Effective Indicators And Assessments For Teaching Ballroom Dance In A High School Setting, Alyssa Arnold

Master's Theses

The purpose of this study was to determine the most effective learning indicators of ballroom dancing for high school dance educators in Utah’s DSD so that a later date, updated rubrics and assessments could be created and presented as artifacts to DSD’s Fine Arts Supervisor for future teachers. The researcher sought to answer the following questions:

Q1 How are teachers currently assessing students’ ballroom dancing?

Q2 What are the most important learning indicators to teachers?

Q3 Are the learning indicators in alignment with the required standards?

This research took place over two sessions held during students’ regularly held ballroom dance …

Navigating The Terrain: Sense Of Belonging In Geology Field Camps, Ezra J. Hammersfeld

Master's Theses

Social success and feelings of belonging within academic settings play critical roles in students’ continued engagement. Students who feel like they belong are more likely to perform better academically, create feelings of academic self-efficacy, and be more satisfied with their educational experiences (Aelenei et al. 2020; Baker 2013; Chen et al. 2020; Hoffman et al. 2021; Rosenthal et al. 2013; Smerdon 2002; Walton et al. 2012). I examine sense of belonging through analysis of secondary data focused on college student experiences in geology field camps. I approached the study using modified grounded theory (Charmaz 2014) in order to best capture …

Integrating Dance Into Early Childhood Education For Children With Developmental Disabilities: A Curriculum-Based Approach, Stephanie Stine

Master's Theses

The purpose of this thesis was to examine the effects of a structured dance curriculum on the social-emotional learning (SEL) and gross motor development of preschool-aged children with disabilities, specifically those ages three to five, as well as to explore teacher and specialist perceptions of dance as a tool for supporting children’s development. The study was conducted in collaboration with an early childhood education center. Twenty-eight children, ages three to five, participated in ten creative movement sessions designed to promote coordination, balance, and expressive engagement through themed lessons emphasizing emotional awareness and imaginative play.

Quantitative data were collected through observation …

Dancing With Logic: The Impact Of Integrating Dance In Teaching Introductory Computer Science Concepts On Student Understanding And Engagement, April Monk

Master's Theses

The purpose of this study was to investigate the effectiveness of dance-integrated pedagogical methods in enhancing the learning experiences of students in an introductory computer science course. In particular, the research aimed to measure the impact of creative movement on student comprehension, engagement, and perceptions of computer science. To guide this investigation, the study explored three research questions: Does integrating dance into lessons impact students’ comprehension of fundamental computer science concepts? How does dance integration affect student perceptions of both dance and computer science? And what elements of dance contribute to differences in comprehension between dance-integrated and traditional lessons? A …

Development And Test Of An Automatic Mass Balancing System For Cal Poly's Spacecraft Attitude Dynamics Simulator, Cameron B. Zorio

Master's Theses

The Cal Poly Spacecraft Attitude Dynamics Simulator (SADS) is an ongoing project to develop an air-bearing platform capable of recreating on-orbit rotational dynamics with near frictionless and torque-free rotations. However, any offset between the platform's center of rotation and its center of mass will introduce a torque due to gravity.

This thesis presents the design, implementation, and experimental evaluation of a low-cost, automatic mass balancing system for the SADS to address this challenge. A new modular sliding mass system was developed that overcomes issues faced in previous iterations of the SADS, providing real-time positional control of the masses and a …

Marsanywhere: Dataset And Cross-View Diffusion Model For Satellite-To-Ground View Synthesis With Mars Data, Benjamin T. Hinchliff

Master's Theses

Satellite-to-ground view synthesis aims to create a realistic ground view image from a corresponding satellite view image. This is a well-studied problem for street level imagery, with good results being achieved by using modern image synthesis techniques such as diffusion models. However, despite the public availability of satellite and ground level imagery on Mars, these techniques have yet to be applied to the domain due to difficulties in collating and processing the data into a usable form. We address this deficiency by creating a dataset consisting of ground view panorama imagery from the Perseverance rover, along with associated satellite view …

Why The Long Face? The Distribution And Niche Of The Imperiled Narrow-Faced Kangaroo Rat (Dipodomys Venustus), Garrett L. Gimbel

Master's Theses

Distribution models provide a practical use case of connecting ecological niche theory with species’ occurrence and assemblages. By explicitly linking Hutchinson’s “niche space” with geographic space researchers can gain valuable information on a species’ niche characteristics and interactions, thereby informing conservation and management practices. The narrow-faced kangaroo rat (Dipodomys venustus) is a highly understudied kangaroo rat species endemic to the central coast of California and is an ideal species to explore niche theory in distribution modeling. Three subspecies are currently recognized: The Santa Cruz kangaroo rat (D. v. venustus), the elephant-eared kangaroo rat (D. v. …

Using Hydraulic And Hydrologic Modeling To Determine Scour At Toro Creek Under Climate Change Conditions, Gabriel E. Alaniz

Master's Theses

ABSTRACT

This study evaluates bridge scour at the Toro Creek crossing of Highway 1 in Morro Bay, California, under present conditions and future climate conditions with more frequent extreme rainfall events. California infrastructure is becoming increasingly vulnerable to extreme storm events and increased intense precipitation events brought by climate change. This analysis focuses on the calculation of scour for both the northbound and southbound spans. Highway 1 plays a vital role in California’s coastal transportation infrastructure by carrying more than 20,000 vehicles daily (Caltrans, 2017). In this thesis, a hydrological model (HEC-HMS) is used to simulate runoff under present-day and …

The Edge Of The World: An Examination Of The Wild West In California, 1840-1880, Chloe J. Cushing

Master's Theses

“The Edge of the World” revolves around the history of the Wild West in California through the lens of Richard Maxwell Brown’s Western War of Incorporation theory. Rather than being a mythology sensationalized by Hollywood and literature, for the Golden State, the Wild West was a very real and tangible part of its early-American era, characterized by U.S. imperialism, high murder rates, anti-authoritarianism, and other rampant forms of chaos that affected the state for generations. By examining the conditions and effects of the late Mexican period, the Gold Rush, Confederate sympathizers, and the Southern Pacific Railroad, this thesis highlights the …

Enhancing Ad/Adrd Management Through Ihelpcare: A Compliant And Culturally Sensitive Ai-Driven Digital Healthcare Platform, Trisha Bhowmick

Master's Theses

The digital healthcare field is expanding fast, and now it requires platforms that use advanced technology and maintain robust data security and compliance practices. In the present paper, we present the main structure, key methods, and compliance strategies of the digital healthcare system iHelpCare, which, while fully meeting the HIPAA/GDPR requirements, provides health services more accessible, efficient, and inclusive. The proposed platform is powered by AI for personalized care solutions, with the main emphasis on preventive health management and providing tools for people with disabilities.

iHelpCare achieves real-time patient monitoring while securing medical data management and easy communication between patients, …

Llm-Powered Question Answering For Object States In Virtual Reality, Shiyi Ding

Master's Theses

Recent advances in large language models (LLMs) and multimodal large language models (MLLMs) enable natural language–based querying in virtual reality (VR). However, VR environments are highly localized, personalized, and dynamic, making it challenging for general-purpose models to answer environment-specific queries or reason about subtle object state changes. To address these challenges, this thesis develops two systems for 3D question answering in VR.

First, we present RAG-VR, the first retrieval-augmented 3D question-answering system designed for VR. RAG-VR augments an LLM with external knowledge retrieved from a localized knowledge database and includes a pipeline for extracting environmental and user-related information. To improve …

Computational Study On Protein Docking With Vascular Disease Application, Lohitha Reddy Shanigaram

Master's Theses

This thesis investigates how computational methods can help us understand protein docking and vascular diseases by integrating machine learning. The main goal of this research is to develop a system that can predict whether variants in the SOX17 gene are associated to Pulmonary Arterial Hypertension (PAH), a significant vascular condition. To do this, a dataset of 1,063 SOX17 variants was created using publicly available sources such as ClinVar, gnomAD, LOVD, and Ensembl. The study included cleaning and structuring the dataset, encoding genomic data and training a Random Forest classifier to classify variants as positive, negative, or unknown for disease association. …

Personality And Perceived Police Legitimacy, Yuxuan Yang

Master's Theses

Police legitimacy refers to one’s trust in the police to be fair and make just decisions in their interactions with the public. Individuals’ perceptions of police legitimacy play a crucial role in predicting how they interact with police. More positive perceptions are associated with cooperative behaviors; less favorable views are associated with resisting behaviors. Few studies have examined the relationship of broad personality traits with perceptions of police legitimacy. The current study examined the HEXACO model of personality and right-wing authoritarianism as predictors of perceived police legitimacy among college students (N = 268). While taking gender, race, and social …

A Study On Deepsea Fish Detection Using Convolutional Neural Networks, Poojitha Priyadarshini Madari

Master's Theses

This study investigates automated deep-learning methodologies for two primary tasks, firstly fish and its habitat classification and deep-sea fish detection using the DeepFish dataset. A pretrained ResNet-50 model attained a validation accuracy of 82.8% for multi-class habitat classification, demonstrating robust performance in most habitats based on 6,517 images, although exhibiting lower recall for visually similar or underrepresented classes. A binary ResNet-50 classifier achieved 99.8% accuracy in distinguishing fish from. no-fish images using approximately 40,000 images. The YOLOv5s model, trained on 4,505 images containing 15,463 bounding-box annotations, achieved a mean average precision ([email protected]) of 98.2% All models were trained on the …

Life History And Distribution Of Two Species Of Cownose Ray, Rhinoptera Bonasus And Rhinoptera Brasiliensis, In Mississippi Coastal Waters, Anna Millender

Master's Theses

Historically, the American Cownose Ray (Rhinoptera bonasus) was considered the only rhinopterid residing in the northern Gulf of Mexico (GOM); however, recent studies have confirmed the presence of the Brazilian Cownose Ray (Rhinoptera brasiliensis) throughout this region. The current study investigated the life history and distribution of both species within Mississippi waters, an area that previously represented a data gap for both species. From April 2016 – July 2025, 193 R. bonasus and 94 R. brasiliensis were analyzed to describe the reproductive biology. Reproductive timing, mating, gestation length, and parturition were similar among species. Although more …

Best Practices For Local Law Enforcement In Counterterrorism: Enhancing Interagency Communication & Developing Modern Information Sharing Systems, Michael G. Fitterman

Master's Theses

Local law enforcement agencies play a critical and increasingly central role in the United States’ domestic counterterrorism architecture. While federal agencies possess sophisticated intelligence capabilities, local officers are often the first to encounter early indicators of radicalization, pre-operational planning, and suspicious behavior. Despite this frontline role, communication between local, state, and federal agencies remains fragmented, inconsistent, and sometimes informal. These gaps weaken national situational awareness and limit the ability of agencies to identify emerging threats.

This thesis examines the structural, technological and cultural challenges hat impede effective information sharing among law enforcement agencies. It outlines best practices that local departments …

Short-Arc Angles-Only Initial Orbit Determination For Leo And Geo Objects Using A Genetic Algorithm, Joseph R. Piini

Master's Theses

An ever-growing interest in utilizing space for scientific, commercial, and defense applications has led to an exponential rise in orbital debris recent years. With this sharp increase in space objects comes a greatly increased risk of potential collisions, necessitating an increase in Space Situational Awareness. In order to best mitigate the risk of future collisions in an increasingly congested space environment, it is crucial to be able to accurately determine the orbits of these new objects for trajectory tracking. This process is initial orbit determination (IOD), and is especially difficult when only observation angles are available for short observation times. …

Forecasting Floral Demand To Reduce Operational Cost And Optimize Production, Mackenzi Saleh Mehlberg-Smith

Master's Theses

The floral industry faces rising operational costs, increasingly strict environmental regulations, and high market volatility, all of which threaten the sustainability of flower production. Accurate demand forecasting and production planning are therefore critical to minimize waste and reduce costs while maintaining competitiveness.

This thesis develops mathematical models for optimization, forecasting, and production planning in the floral industry, extending recent work on production planning and trapezoidal demand modeling. We integrate time-series forecasting methods with optimization techniques from production planning and operations research to model both short-term demand fluctuations and medium-term capacity constraints. The mathematical framework links linear and nonlinear optimization, time-series …