Digital Commons Network^™

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

Theses

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

Heathammer: Effects Of Thermal Stress On Dram Technology Reliability Using Rowpress And Rowhammer, Filip Roth Tronnes-Christensen

Theses

Modern DRAM scaling has reduced cell capacitance and increased thermal sensitivity, making disturbance-based faults such as RowHammer and RowPress increasingly significant reliability and security concerns. RowHammer induces bit flips through repeated row activations, while RowPress does so by holding a wordline open for an extended duration; both exploit inherent capacitive coupling and leakage mechanisms in dense DRAM arrays. This thesis introduces HeatHammer, a thermally assisted disturbance exploit that interleaves RowPress and RowHammer operations to amplify charge leakage and trigger row-traversing bit flips. Using the FPGA-based DRAM-Bender test platform, HeatHammer is evaluated on four commercially available DDR4 modules from different manufacturers …

A Study Of Perceptions, Readiness, Benefits, And Barriers Related To Exoskeleton Adoption In New Jersey’S Warehousing Sector, Terry Asante

Theses

The warehousing industry in New Jersey remains a vital component of the region's logistics network, employing more than 200,000 workers who routinely engage in lifting, bending, overhead reaching, and other physically demanding activities. These exposures contribute to musculoskeletal disorder (MSD) rates that exceed national averages, particularly affecting the low back and shoulders. Nationally, MSDs account for an estimated $420 billion in combined direct and indirect costs each year, underscoring the need for interventions that can effectively reduce biomechanical strain. Industrial exoskeletons have emerged as a potential solution, with prior research demonstrating reductions in muscle activation, perceived exertion, and fatigue during …

Simulation-Based Assessment Of Operational Strategies For Visibility-Responsive Variable Speed Limits, Jiyeon Ban

Theses

Managing speed limits in adverse weather is crucial for highway safety. Variable Speed Limit systems aim to adjust speed limits according to weather conditions, yet many traffic management centers depend on manual operator judgment. Rapid visibility fluctuations, along with operators handling multiple responsibilities, can cause delays in responding to deteriorating conditions, increasing the risk of rear-end collisions. A VSL control strategy is developed using real-time Road Weather Information System data to enable more proactive and effective adjustments. Microsimulation analysis indicates that visibility-responsive VSL operations can reduce conflicts by 59% in fog. Applying VSL only to fog-affected sections yields clear safety …

Evaluating Stormwater & Carbon Benefits Of Urban Street Trees, Anthony J. Rodriguez Diaz

Theses

Urban areas experience a multitude of environmental challenges, including flooding and air pollution. In response to these pressures, urban forests have emerged as an essential nature-based strategy to enhance resilience and reduce climate-related risks. The implementation and expansion of urban street trees further contribute to this effort by mitigating stormwater runoff and reducing atmospheric carbon emissions.

Fortunately, advances in environmental modeling have made it increasingly possible to quantify these ecosystem services using specialized software. The United States Department of Agriculture (USDA) Forest Service's i-Tree suite exemplifies this progress. The software provides a comprehensive set of tools designed to quantify environmental …

A Novel Rheological Technique To Measure The Yield Stress In Equibiaxial Elongation, Asher Segal

Theses

Yield stress measurement in non-Newtonian fluids, particularly under complex flow conditions such as equibiaxial elongation, remains a significant challenge in experimental rheology. This work presents the development of a novel method, Continuous Lubricated Squeezing Flow (CLSF), designed to quantify the normal yield stress of viscoelastic materials while minimizing boundary artifacts.

The CLSF method was implemented and validated using Carbopol 940, a model yield-stress gel, at concentrations of 1 wt% and 2 wt%. Comparative shear rheology was first performed to establish baseline yield stresses and verify sample integrity. CLSF experiments were then conducted over controlled flow rate increments and gap distances …

Optimizing Crime Response Time Using Predictive Analytics Models, Mohammad Hassan Bin Yaqoob

Theses

Timely and efficient emergency response is essential for public safety, particularly in high- risk urban environments. However, existing dispatch systems often rely on reactive decision- making, leading to delays and inefficient resource allocation. This thesis explores how artificial intelligence (AI) and data modeling can enhance police response times through predictive analytics. A systematic review of 17 peer-reviewed studies identified key predictors such as dispatch behavior, crime history, temporal characteristics, and geographic location. Using real-world emergency response and crime data, predictive models were developed and evaluated through algorithms including XGBoost, Linear Support Vector Machine (LSVM), Generalized Linear Model (GLM), and Neural …

A Fish & A Bird, Deanna Moorehead

Theses

As a curtain of trees opens and dawn breaks over the glistening waters of Lake Ontario, a fish and a bird spot each other through the surface and fall madly in love. They dance and sing together from their respective atmospheres, yet each time the bird soars past the water's edge and onto the beach, the fish cannot follow. Night after night, by the light of the moon, the fish trains herself to leap out of the water, hoping to breach high enough to join her beau above. Late one afternoon, the fish approaches the shore, where the bird stands …

Scout 07, Austin Roberts

Theses

Throughout my tenure at RIT, I have produced multiple short films, undertaken gaming co-ops, and pursued an independent study centered on gaming assets to expand my technical and creative skills. For my thesis film, I sought to merge these experiences by concentrating on environmental storytelling—a deliberate challenge, given faculty concerns that narratives lacking strong character focus may struggle to engage audiences emotionally. With these considerations in mind, I crafted a film featuring a secondary character within a rich environment, designed to immerse viewers and provoke curiosity about the story’s temporal and spatial context.

Evaluating The Feasibility Of Digital Twin Adoption For Mep Systems, Tejas Adana

Theses

MEP systems dominate building operational costs, yet facility teams continue battling reactive maintenance cycles that result in unexpected equipment failures, emergency callouts, and continuous energy inefficiencies. Digital twin technology presents a practical solution by creating live virtual replicas that shift operations from firefighting breakdowns to anticipating issues before they escalate. However, adoption remains limited due to fragmented implementation guidance, short-term validation studies, and minimal research addressing emerging market contexts. This research employs a grounded approach combining systematic literature analysis, candid discussions with MEP engineers and facility managers across UAE commercial properties, and practical prioritization frameworks assessing real-world impact against implementation …

Disposal, Yihan Su

Theses

Disposal is a mixed-media animated short film created as a graduate thesis project. The story explores the protagonist’s search for meaning in life and his defiance of a predetermined fate. At the beginning, he is an ordinary, guileless figure without remarkable qualities. In his world, life revolves around constant competition, and one’s worth is measured by their position on a ranking board. Only the top three may pass through the golden gate, a symbol of transcendence. Others must continue the endless cycle of contests, while the lowest seven face incineration in the Disposal. Despite his efforts, the protagonist falls into …

Hugs In Hues, Shivani Pednekar

Theses

Hugs in Hues is an animated short film about social anxiety, emotional expression, and discomfort with physical touch. The story follows Tina, a young artist who struggles to express how she feels. When she pulls away from a hug from her friend Emily, Tina escapes into a painting-like 2D world made from her drawings. This imagined world reflects Tina’s emotions. The colors, textures, and spaces change based on how she feels. Her anxiety, fear of rejection, and need for safety take visual form. The real world feels sharp and rigid, while Tina’s illustrated world feels soft, imperfect, and expressive. Animation …

The Impact Of Screen Time Patterns On Digital Behaviour And Productivity: A Data Analytics Study Across Age Groups, Maytha Alshamsi

Theses

The increasing incidence of screen time of all ages has generated a greater interest in the effect of screens on the human psyche. This paper looks at behavioural data of 2, 000 individuals aged 13-64 to investigate the predictability of a composite mental- health score by various modalities of screen-time, lifestyle and psychological charac- teristics. We determine, based on exploratory statistics, multivariate regression and su- pervised machine-learned models, that in as much as the influence of total screen time and social-media consumption has significant relationships with worse mental health, the effects are small when juxtaposed to those of lifestyle and …

Content Aware Intra And Inter Prediction Using Texture And Motion Analysis, Yashaswi Karthik Reddy Danda

Theses

Modern video codecs rely on highly optimized block based intra and inter prediction frameworks,  in which block partitioning is determined through exhaustive rate distortion optimization (RDO)  searches. These exhaustive searches evaluate many possible block configurations to identify the  partitioning that minimizes overall rate distortion cost, and they have proven extremely effective  across diverse visual content. This thesis explores a complementary perspective by investigating  whether block partitioning can be guided more directly by content characteristics specifically,  texture complexity for intra prediction and motion complexity or motion density for inter  prediction.  For intra prediction, we introduce a texture adaptive approach that utilizes …

The Impact Of Artificial Intelligence On Employee Attrition And Organizational Profitability, Saeed Almehairi

Theses

Artificial Intelligence (AI) has become a transformative force in the modern workplace, reshap- ing decision-making processes, enhancing operational efficiency, and influencing the dynamics of human resource management. One of the critical areas where AI demonstrates substantial potential is in predicting, managing, and mitigating employee attrition. Employee attrition remains a pressing organizational challenge, incurring substantial costs in recruitment, training, and lost productivity. This research investigates the dual impact of AI on employee attrition and organizational profitability, exploring both the opportunities and the challenges that arise from its integration into human resource and business operations. The study examines how AI-driven predictive analytics …

Planning For Power: The Role Of Architecture & Urban Form In Local Energy Generation In Dense Built-Up Areas, Chris Shing Wai Leung

Theses

This thesis explores how the urban form/building design, and a well-thought-out City Master Plan can significantly affect wind patterns and enhance on-site energy generation in dense urban areas. It will examine strategies such as modifying building heights, adjusting separations between structures, shaping urban landforms, and varying the roughness of the urban fabric to optimize wind energy generation. Additionally, the research will focus on the strategic placement of community power harvesting devices. The study of the urban forms and fabric with the impact of wind patterns is simulated by fluid mechanics with Computational Fluid Dynamics (CFD) software. The goal of this …

Basil_Dock: Development Of Accessible And Customizable Molecular Docking Procedures, Lee Schoneman

Theses

Molecular docking is a computational technique often used in drug design due to its ability to predict ligand binding potential, conformation, and location for a given drug target. Docking as a screening tool also comes at a greatly reduced cost compared to wet bench work. Despite its usefulness, molecular docking is often inaccessible to novice users, as to perform molecular docking, most programs require at least a basic knowledge of command line and computer programming for both installation and utilization. Additionally, tutorials for the most popular programs tend to be inflexible, instructing how to bind predetermined molecules to a specified …

A Decision Support System For Global Vaccine Funding: Data-Driven Proposal Scoring, Epidemiological Risk Modeling, And Portfolio Optimization, Ashutosh Kumar

Theses

Global health financing must continually stretch limited immunization resources across competing priorities, and Gavi, the Vaccine Alliance, plays a central role in this landscape. This work reframes Gavi’s funding decisions as a multi-objective portfolio optimization problem that jointly considers health impact, equity, cost-effectiveness, sustainability, and epidemiological interdependence. To operationalize this approach, this study develops a decision support system grounded in Modern Portfolio Theory, treating each proposal as an asset characterized by expected return and risk. Proposal returns are estimated through a heterogeneous data-fusion pipeline that generates quantitative, multi-objective scores aligned with Gavi’s strategic priorities, while epidemiological risk is quantified using …

Comparative Analysis Of Machine Learning Models For Spam Email Detection, Rashed Almarri

Theses

The current research examines one of the most effective approaches to spam email detection based on machine learning and natural language processing (NLP). The study is placed in the context of the rising cyber threats and the influx of emails, where the spam/ham data is to be classified correctfully using the combination of the Logistic Regression, NLP (including tokenization, lemmatization, and TF-IDF vectorization). The questions of the research were devoted to the efficiency of such an approach and the interpretation of its results. The data used are obtained by a publicly available Kaggle data set that contains 5,572 labeled email …

Security Vulnerabilities In Cloud Storage: A Comparative Study Of Google Drive, Dropbox, And Onedrive, Sultan Majid Alshamsi

Theses

In recent years, use of internet has increased manifold. People have started creating more data, and instead of saving it locally, people have started preferring storing it “online”. Companies are also following the same path to maintain accessibility and availability of their data. Thus, data has become a central part of our lives. People, companies, and institutions rely heavily on cloud storage solutions these days for managing their data. Some of the main cloud storage solution are Google Drive, Dropbox, and OneDrive. These are sophisticated solutions developed by tech giants, and general expectation of the public is that these solutions …

Geometric Deviation Prediction In Additively Manufactured Parts Using Conditional Generative Adversarial Networks Under Varying L-Pbf Process Parameters, Subigyamani Bhandari

Theses

The progress in metal AM technology has enabled printing parts with thin features and intricate geometries. Predicting and reducing metal AM part geometrical deviations to narrow down the difference between printed and designed structure still poses a problem. This work explores changes in laser speed, laser power, and hatch spacing affecting geometrical deviations in parts made using laser powder bed fusion (L-PBF) and places an emphasis on how to predict geometrical defects in the AM parts. Sliced images obtained from CAD designs and printed parts to capture the effect of various L-PBF process parameters are utilized to create a dataset. …

Comprehensive Drought Analysis: Machine Learning-Based Meterological Drought Forecasting And Pca-Driven Agricultural Drought Monitoring., Bishal Poudel

Theses

Drought is a long-term natural disaster that affects many aspects of human life from health, water supply to ecosystems and agriculture. Drought’s occurrence and intensity has increased in recent years because of global warming, which urges for proper and accurate monitoring and forecasting of drought. For better understanding of droughts, this study uses two approaches. In first part, the study uses the historical temperature and precipitation data from period of 1960 to 2021 as input features for three different machine learning models – Artificial Neural Network (ANN), Support Vector Machine (SVM) and Random Forest (RF). The research focuses on calculating …

Prediction Of Geometric Deviation In Additively Manufactured Tpms Lattice Structures Using Non-Destructive Evaluation And Conditional Generative Adversarial Networks, Prateek Neupane

Theses

Metal 3D printing has emerged as a widely used additive manufacturing method, finding applications across numerous industries and research areas. However, its application is often hindered by geometric inaccuracies and surface quality issues resulting from the complex interplay of laser power, melt pool behavior, and cooling characteristics. These inconsistencies have a significant impact on the design and functionality of printed parts, affecting structural integrity and dimensional accuracy. Thus, an effective data-driven approach is crucial to predict these deviations before manufacturing. This study proposes a methodology that incorporates Non-destructive evaluation techniques, like in-situ monitoring and X-ray CT scan, with Conditional Generative …

Exploring The Adoption Of Quality 4.0: A Qualitative Study Of Industry Practices And Barriers, Tama Rani Sarker

Theses

The emergence of Industry 4.0 has revolutionized quality management, introducing Quality 4.0 as a framework that integrates digital technologies with established quality principles. Despite its potential to enhance operational efficiency and decision-making, many organizations face challenges in adopting Quality 4.0, including skill gaps, resource constraints, and cultural resistance. This qualitative study explores industry practices, barriers, and enablers of Quality 4.0 implementation through surveys and interviews with quality professionals in southern Illinois manufacturing companies. Using a phenomenological approach, five major themes emerged: employee skillset, training, technology integration, quality tools, and supply chain management. Findings reveal that successful adoption depends on workforce …

Effect Of Water-Deficit Stress On Cannabis Sativa Production And Secondary Metabolite Levels, Shiksha Sharma

Theses

Cannabis (Cannabis sativa L.) is an emerging high-value crop that belongs to the Cannabaceae family and the genus Cannabis, which is known to produce more than 200 cannabinoids. Although genetic variation is the main factor in cannabinoid production, water-deficit stress is believed to induce its production. The objective of this master’s thesis was to determine the effects of water-deficit stress frequencies and timing on growth, physiology, yield, and cannabinoid concentration in cannabis. Heidi cultivars were planted in a controlled environmental growth unit. One period of water-deficit stress was found to produce the lowest concentration of total CBD and THC. In …

Flood Susceptibility Assessment Through Gis Integrated Analytical Hierarchy Process And Machine Learning Models, Sujan Shrestha

Theses

Flooding is among the most destructive natural disasters globally, and it inflicts severe damage on both natural environments and human-made structures. The frequency of floods has been increasing due to unplanned urbanization, climate change, and changes in land use. Flood susceptibility maps help identify at-risk areas, supporting informed decisions in disaster preparedness, risk management, and mitigation. This study aims to generate a flood susceptibility map for two regions: Davidson County of Tennessee using an integrated geographic information system (GIS) and analytical hierarchical process (AHP), and the Briar Creek watershed of Mecklenburg County, North Carolina using an integrated GIS and machine …

Asymptote, Aybuke Yilmazer

Theses

Asymptote is a short film that is an expression of mankind’s emotional consumption through confusing love with sex. This film is a frame-by-frame hand-drawn animation that combines digital and analog techniques. It was produced between August 2024 and April 2025, and its music was re-produced in August 2025. This paper is a written declaration of the motives and intentions of the director, Aybuke Yilmazer, in making this film, and analyzes the filmmaking process. The text of the thesis is in third person except for the self-evaluation section.

Accessible Preservation: Developing A Low-Cost, Iiif-Compliant Workflow For Small Art Collections, Robyn Palescandolo

Theses

This project explores the development of an affordable, standards-based digital collections management system designed for small galleries, independent artists, and emerging institutions. Using the artist’s own studio collection as a case study, the research integrates accessible hardware and open-source technologies, including the International Image Interoperability Framework (IIIF), Artwork Archive, GitHub Pages, and the Internet Archive, to establish a transparent, replicable workflow for cataloguing and publication. The project demonstrates how interoperability, metadata consistency, and open access can be achieved without reliance on costly proprietary systems, such as TMS Collections or PastPerfect.

The workflow includes high-resolution image capture, 3D scanning, data normalization, …

Advancing 3d Infrastructure Utilities Mapping Through The Integration Of Multi-Method Digital Modelling Techniques, Adil Abdulla Aldashti

Theses

With urban growth accelerating, cities increasingly need precise 3D information about their underground service networks to support sustainability goals and digital systems. Older 2D maps are often outdated, incomplete, and misaligned, making them difficult to use within 3D GIS environments. Moreover, conventional surveying tools may fall short in accuracy, struggle to reach buried assets, and require high costs when dealing with existing utilities. This thesis aims to develop and evaluate a field to 3D methodology for mapping underground utilities in Dubai, integrating multi-sensor and surveying approaches, and GIS 3D models in alignment with ASCE 38-22(American Society of Civil, 2022a), PAS …

Machine Learning Approaches For Predicting And Preventing Vehicle Theft, Hamad Almutawa

Theses

This study develops a data-driven framework for predicting vehicle theft using machine-learning techniques applied to a comprehensive crime dataset containing demographic, spatial ,and operational policing variables. The dataset, sourced from publicly available crime records, included 9,712 cases and 30 variables, with minimal missing data and balanced class distribution. After performing extensive preprocessing—including outlier analysis, missing-value imputation, categorical encoding, feature selection, and correlation testing—four supervised classification models were implemented: Logistic Regression, Random Forest, Support Vector Machine, and a Neural Network. Bivariate statistical tests, including Chi-square and Mann–Whitney U, revealed significant relationships between theft occurrence and factors such as crime area, luxury …