Digital Commons Network^™

Ransomware As Organization: A Comparative Analysis Of Corporate And Criminal Structures In Conti, George Urling

Theses, Dissertations and Capstones

Cybercriminal groups continue to pose major threats to global cybersecurity. One of the most common types of cybercriminal groups are, “Ransomware-as-a-Service (RaaS)" groups, who create and sell ransomware. While research is conducted into the development of ransomware, there is limited reporting on the organizational structure and habits of RaaS groups. In 2022, prominent RaaS group Conti had their chat logs leaked, with the logs ranging from 2020 to 2022. This study seeks to provide a deeper understanding of RaaS group structures by utilizing the Conti leaked logs as a case study. The study, entitled “Ransomware as Organization: A Comparative Analysis …

A Macrocognitive Design Taxonomy For Simulation-Based Training Systems: Bridging Cognitive Theory And Human-Computer Interaction, Jessica M. Johnson

Virginia Digital Maritime Center (VDMC) Faculty Publications

Simulation-based training systems are increasingly deployed to prepare learners for complex, safety-critical, and dynamic work environments. While advances in computing have enabled immersive and data-rich simulations, many systems remain optimized for procedural accuracy and surface-level task performance rather than the macrocognitive processes that underpin adaptive expertise. Macrocognition encompasses higher-order cognitive processes that are essential for performance transfer beyond controlled training conditions. When these processes are insufficiently supported, training systems risk fostering brittle strategies and negative training effects. This paper introduces a macrocognitive design taxonomy for simulation-based training systems derived from a large-scale meta-analysis examining the transfer of macrocognitive skills from …

Towards Multimodal Guideline-Aligned Agentic Systems, Wenliang Zhong

Computer Science and Engineering Dissertations - Archive

I present my work on building multimodal guideline-aligned agentic systems designed to enable AI agents to solve complex real-world tasks. My research addresses two critical perspectives: (1) Instruction-Aware Embedding Models for flexible and universal embedding tasks, and (2) Guideline-Driven LLM Agents that leverage domain-specific guidelines to perform expert-level tasks. These components address embedding and generation tasks, respectively, and lay the foundation for a hybrid agent capable of tackling challenging real-world applications.

From the embedding perspective, I first address the instruction-following capabilities of embedding models. While Large Language Models (LLMs) excel at instruction following, they are primarily designed for generation rather …

Computational Clinical Judgment: Predicting Risk With Large Language Models, Hannah Laqueur, Ryan W. Copus

Faculty Works

For seventy years, research has shown actuarial methods outperform clinical judgment. Yet actuarial approaches have limitations: they generally rely on structured data; cannot exploit rare case-specific details; have limited accuracy where outcome data are scarce or incomplete; and cannot offer case-level justifications. Large language models (LLMs) offer a different approach. Like actuarial methods, they aggregate information algorithmically, but like clinicians, they bring general knowledge and can provide case-level justifications. We prompted seven LLMs to assess rearrest risk from 113 parole hearing transcripts and compared their predictions to a machine learning model trained on 4,000 cases with 91 administrative variables. GPT-5 …

Do Algorithms Dream Of Electric Muses? Teaching Creative Writing In The Age Of Generative Ai, Anna Leahy

English Faculty Articles and Research

In Philip K. Dick’s novel Do Androids Dream of Electric Sheep? androids are given a psychological test to confirm they are not human before killing them. The story’s end suggests that humans will treat a seemingly harmless android as authentically as a human even when humans are aware the android is not human. Students use tools like ChatGPT, which function as autocomplete on steroids, to produce text using probabilistic relationships among words, and instructors can’t always tell the difference between average student writing and Gen AI text. In creative writing classes, instructors might use thinking for oneself as a central …

Multimodal Deep Learning For Biological Data Understanding, Saiyang Na

Computer Science and Engineering Dissertations

This dissertation presents three contributions to multimodal deep learning for biological data understanding, addressing the fundamental challenge of cross-modal alignment from two complementary perspectives: designing effective multimodal fusion methods for specific biomedical applications, and proposing a general framework for higher-order multimodal alignment that captures hierarchical structure in data.

First, we develop Cmai, a deep learning framework for B cell receptor (BCR) to antigen binding prediction that aligns BCR sequence information with antigen three-dimensional structures using contrastive learning. Cmai achieves an average AUROC of 0.907 across 17 antigens and 5 independent cohorts, and demonstrates clinical utility in predicting immune checkpoint inhibitor …

A Hybrid Response Surface Methodology And Machine Learning Framework For Quantifying Effects Of Physicochemical Parameters On Pfas Distribution, Harsh V. Patel, Jazmin Green, Hyoshin Park, Stephanie Luster-Teasley Pass, Renzun Zhao

Engineering Management & Systems Engineering Faculty Publications

Predicting PFAS adsorption across diverse adsorbents and environmental matrices remains challenging because adsorbent physicochemical properties, PFAS molecular descriptors, and operational conditions simultaneously influence adsorption. This study develops and evaluates a unified hybrid modeling framework that integrates Response Surface Model (RSM) with machine-learning algorithms to quantify how six key variables, surface area, Log K_ow, pH_pzc, pK_a, log dose, and log-initial concentration, affect PFAS distribution coefficients (Log K_d). A data set of more than 1000 adsorption observations spanning 15 PFAS compounds, multiple adsorbent types, and a broad operational range was compiled and preprocessed using …

Ai Deployment Authorisation: A Global Standard For Machine-Readable Governance Of High-Risk Artificial Intelligence, Daniel Djan Saparning

Student Publications

Modern artificial intelligence (AI) governance lacks a formal, enforceable mechanism for determining whether a given AI system is legally permitted to operate in a specific domain and jurisdiction. Existing approaches-such as model cards, audits, and benchmark evaluations provide descriptive information about model behaviour and training data but do not produce binding deployment decisions with legal or financial force. This paper introduces the AI Deployment Authorisation Score (ADAS). This machine-readable, regulator-grade framework evaluates AI systems across five legally and economically grounded dimensions: Risk, Alignment, Externality, Control, and Auditability, derived from safety engineering, alignment theory, algorithmic accountability, and liability economics. ADAS produces …

Detecting And Repairing Conflicting Constraints In Co-Trained Physics-Informed Neural Networks For Composite Curing Processes, Cooper J. Evans

Dissertations, Master's Theses and Master's Reports

Composite materials have become a critical component of modern manufacturing, especially in the automotive and aerospace industries. The curing process for these composites has been modeled using a variety of partial differential equations representing the heat transfer and composite curing kinetics. Optimizing the applied temperature profile is critical for maximizing the efficiency and capacity of composite part manufacturers. Constraints must be placed on the inputs and outputs of the model, including but not limited to, the applied temperature profile, part temperature, and final degree of cure. Conflicting sets of constraints are easy to unknowingly impose due to the highly coupled …

Systematic Approaches To Characterizing Vulnerabilities And Enhancing Robustness Of Text And Vision-Language Models, Poojitha Thota

Computer Science and Engineering Dissertations

The proliferation of artificial intelligence (AI) across critical domains, including news summarization, privacy-policy analysis, and medical decision support, has raised growing concerns about the security and robustness of these systems against adversarial manipulation. This dissertation investigates adversarial robustness in generative AI by addressing three key research goals: (1) characterizing adversarial vulnerabilities across generative models, (2) developing systematic defenses to improve the robustness of generative models, and (3) designing deployment-time safeguards for securing LLM interactions.

Towards the first goal, we characterize adversarial vulnerabilities across text-based and multimodal systems. In abstractive text summarization, we show that inference-time perturbations can exploit lead bias …

Beyond Fixed Thresholds: Per-Label Calibration For Fine-Grained Emotion Detection On The Goemotions Dataset, Sai Puneet Naga Venkata Subramanyam Patchipulusu

Selected Full-Text Master Theses 2021-

This study investigates the effectiveness of five community fine-tuned transformer models for fine-grained emotion detection on the GoEmotions dataset: SamLowe/roberta- base-go_emotions (RoBERTa-base), cirimus/modernbert-base-go-emotions (ModernBERT), mrm8488/deberta-v3-base-goemotions (DeBERTa-v3-base), bhadresh-savani/bert-base-go- emotion (BERT-base-cased),and tasinhoque/distilbert-go-emotions (DistilBERT) . While the original GoEmotions research by Demszky et al. (2020) established a BERT-base baseline with a macro-F1 of 0.46, this thesis extends that work through independent empirical evaluation of five derivative models, systematic per-label threshold optimization, and comparative analysis of architectural trade-offs across the full transformer model family. Using the GoEmotions simplified test split (5,427 examples across 28 categories), all five models were evaluated at a fixed 0.5 …

Vision-Language System Using Open-Source Llms For Consent And Instruction Gestures In Medical Interpreter Robots, Tung Ngo, Emma Murphy, Robert Ross

Conference papers

Effective communication is vital in healthcare, especially across language barriers, where non-verbal cues and gestures are critical. This paper presents a privacy-preserving vision-language framework for medical interpreter robots that detects specific speech acts (consent and instruction) and generates corresponding robotic gestures. Built on locally deployed open-source models, the system utilizes a Large Language Model (LLM) with few-shot prompting for intent detection. We also introduce a novel dataset of clinical conversations annotated for speech acts and paired with gesture clips. Our identification module achieved 0.90 accuracy, 0.93 weighted precision, and a 0.91 weighted F1-Score. Our approach significantly improves computational efficiency and, …

Regulating Ai Beyond Product Liability, Shruti Trikanad

Michigan Technology Law Review

Artificial Intelligence (AI) is being used by governments across the world to enforce regulatory mandates, adjudicate benefits and privileges, predict and analyze risks, and much more. Although this has significant potential to increase efficiency and responsiveness, it also comes with several risks of transparency, government accountability, and the amplification of discrimination and bias. It is crucial we oversee and regulate these AI systems effectively. This essay argues against the models that current regulatory frameworks are adopting to govern AI use: those resembling product liability.

Through the lens of the European Union's AI Act and Liability Directive, it highlights the unsuitability …

Trustworthy Multimodal Ai For Medical Imaging: Enhancing Diagnosis, Reasoning, And Human-Agent Interaction In Extended Reality, Jai Prakash Veerla

Computer Science and Engineering Dissertations

The transition from traditional microscopy to digital pathology has digitized diagnostic data, yet clinical workflows remain constrained by two-dimensional screens and passive, opaque analysis tools that fail to capture the spatial complexity of biological systems. While Foundation Models now promise to reason across histology and genomics, a critical disconnect persists between the richness of this data and the limited cognitive bandwidth of clinicians, who currently lack the immersive interfaces and trustworthy agents necessary to utilize it effectively. This dissertation presents a unified framework for "Embodied Agentic AI," establishing a pipeline that augments physician capabilities through immersive visualization, robust security, and …

Impact Of Cross-Client Heterogeneity In Federated Learning For Real-World Plant Disease Classification, Ritesh Janga

All Graduate Theses, Dissertations, and Other Capstone Projects

Deep learning applications are being adopted in agricultural image analysis that include challenges of data privacy and limited institutional data and heterogeneity of different types of architectures. However, Federated Learning is a model that allows collaborative training on data that does not have to be shared among parties. Therefore, Federated Learning is an effective method of collaborative training; however, its comparative effectiveness as compared to individual (local) training on diverse architectures has never been examined in an agricultural context. The objective of this study was to examine Federated Learning for the purpose of crop disease classification on extreme non-IID distributed …

Artificial Intelligence Adoption In The Workplace. An Exploration Of Augmentation, Oyinkansola O. Sodiya

Management Dissertations

As collaborative work with artificial intelligence (AI augmentation) gains interest, it is crucial to investigate factors that affect how employees perceive and use AI tools at work. Drawing on task-technology fit and technology adoption theories, this dissertation examines the ways in which task dimensions, organizational contexts, and individual differences affect the perceived usefulness of working with AI tools. This dissertation demonstrates that task-technology fit is fundamental. Employees in jobs with high information processing demands are likely to positively perceive the usefulness of AI augmentation relative to employees in jobs with high interpersonal demands. Employees with more proactive personalities perceive greater …

Toward Interpretable Multi-Omics Multimodal Biomedical Artificial Intelligence, Yanjun Lyu

Computer Science and Engineering Dissertations

The complexity of human disease arises from biological processes that unfold across multiple scales, from molecular variation through cellular function, tissue organisation, brain phenotypes, each of which is associated with distinct measurement modalities, regularities, and characteristic. Contemporary biomedical artificial intelligence has brought the opportunity to reveal the complexity with in; however, its methodological default, in which models are trained on most readily available modality, does not adequately engage with the multi-scale connected structure by which biological meaning is constituted. The research area of multi-omics and multi-modal AI for biomedicine remains at an early exploratory stage, and the work presented in …

Interpretable Machine Learning For Bridge-Pier Scour Prediction And Flood Resilience, Adil Khan, Dalya Ismael

Engineering Technology Faculty Publications

Bridge-pier scour is a leading cause of flood-induced bridge failure, yet practice still lacks transparent, physics-informed tools that link data-driven prediction with design guidance. This study develops an interpretable, physics-aware machine-learning framework to predict equilibrium scour depth and translate those predictions into actionable strategies for flood-resilient infrastructure. Using the 2014 U.S. Geological Survey Pier-Scour Database (569 laboratory cases), five models: Gradient Boosting, AdaBoost (Tree), XGBoost, Gaussian Process (RBF kernel), and Kernel Ridge (polynomial), were trained and evaluated with K-fold cross-validation. Model performance was evaluated using R², RMSE, and MAE. Gradient Boosting performed best, achieving training and testing R² of 0.99 …

Srgan-Based Deep Learning Framework For Wind Turbine Damage Detection From Sentinel-2 Imagery, Kübra Çakir, Onur Elma, Murat Kuzlu

Engineering Technology Faculty Publications

The operational reliability of wind turbines is critical for sustainable energy production in smart grids. This study proposes a remote monitoring approach using perceptually enhanced satellite imagery. Sentinel-2 multispectral data (10 m resolution) has been processed with a Super-Resolution Generative Adversarial Network (SRGAN) to improve visual quality to a perceptual resolution of 30 cm. Although true spatial refinement is not achieved, the sharper structural details enhance classification accuracy. The data set comprises 15,000 images—10,000 SRGAN-enhanced and 5000 augmented through rotation, zoom in, increasing brightness, noise addition, and blurring. A custom Convolutional Neural Network (CNN) has been trained to classify turbines …

Ablative Study Of Large Language Model-Based Gesture Inference For Autonomous Navigation, Neil Loftus

Theses, Dissertations and Capstones

Human gesture inference has broad applications ranging from sign language interpretation to device control. Traditional methods often rely on extensive manually labeled hand datasets for deep learning. Furthermore, they are typically limited to a discrete set of gestures existing in these datasets. Large Language Models (LLMs) created by enterprise companies such as OpenAI have demonstrated positive results in many artificial intelligence tasks, with a notable strength being their adaptability. Existing literature has shown that LLM based systems can not only perform gesture inference but can propose user intent provided with a context and list of possible actions. We propose an …

Neuroevolution Of Compact Search Heuristics In Sokoban, Yeonghun Lee

Computer Science Honors Papers

This thesis investigates whether neuroevolution can produce heuristic functions for A* search on Sokoban that are drastically more parameter-efficient than a conventional gradient-trained baseline. Sokoban is a combinatorial planning puzzle that is both NPhard and PSPACE-complete, and solving it at scale requires an accurate cost-to-go heuristic to guide search. The standard approach fixes a neural network architecture and trains it by gradient descent; this thesis investigates whether evolutionary search over network topology can find heuristics that are competitive in accuracy while using orders of magnitude fewer parameters.

Four heuristic approaches are implemented and compared on the medium-difficulty split of the …

A Novel Lightweight Framework For Low-Light Image Enhancement Via Gaussian Denoising And Clahe, Daniel Oluwaseun Adesoji

Master's Theses or Doctor of Nursing Practice

Low-light image enhancement is a major challenge in digital imaging, especially in medical imaging, surveillance, and autonomous vision systems. Images captured under poor illumination often appear dark, noisy, and low in contrast, which makes it hard to observe important details. Traditional enhancement methods can improve brightness but usually introduce artifacts or increase noise. Although deep learning methods have shown strong performance, they usually require large datasets, and high computational resources. This creates a need for simpler and more efficient enhancement techniques. This study proposes a lightweight framework that incorporates Gaussian denoising with Contrast Limited Adaptive Histogram Equalization (CLAHE) to enhance …

Algorithmic Trading In Idiosyncratic-Payoff Markets: A Multi-Agent System For On-Chain Prediction Contracts, Saif Aldeen A.K. Agha

CMC Senior Theses

This thesis documents the design, deployment, and forward-test evaluation of an evolutionary multi-agent algorithmic trading system on Polymarket, the largest decentralized prediction market. The system pairs a locally-hosted 72-billion-parameter language model with a gradient-boosted statistical filter and an evolutionary selection mechanism that maintains a population of approximately 500 autonomous trading agents. Each agent generates a probability estimate for an event, compares it to the prevailing market price, and trades the resulting disagreement.

The central empirical exercise estimates a panel regression of trade-level profit on the absolute disagreement between the agent's probability estimate and the market price, controlling for agent identity, …

Training Set Augmentation And Biology-Aware Harmonization Improve Radiomic Models For Lung Cancer Prediction In Indeterminate Nodules, Claire Huchthausen, Menglin Shi, Gabriel De Sousa, James Larner, Einsley Janowski, Jonathan Colen, Krishni Wijesooriya

Data Science Faculty Publications

CT radiomics-based machine learning has potential to predict lung cancer in pulmonary nodules (PNs) earlier than standard-of-care methods. Low malignancy rates in early-development PNs and variable image acquisition hinder development of radiomic models for diagnosing these PNs. To address these challenges, we augmented training using later-development PNs and harmonized for acquisition effects. We examine early-development benign and malignant PNs (n = 106) below the sensitivity of standard-of-care diagnosis. Classifiers predicting malignancy performed near chance when trained on ComBat-harmonized radiomic features from only early-development PNs. We then augmented training with later-development benign and malignant PNs (n = 225). We evaluated whether …

Zero-Shot Segmentation Of Estuary Mudflats Using The Segment Anything Model, Jaren Unzen

Honors Theses and Capstones

Estuary mudflats are ecologically sensitive environments that require consistent monitoring. Traditional satellite-based classification workflows are often constrained by the high cost and labor-intensive nature of manual data annotation. This study evaluates the utility of Segment Anything Model 3 (SAM 3), a foundational computer vision model, to automate mudflat segmentation without domain-specific fine-tuning. By leveraging the model’s text-prompting capabilities alongside specialized pre- and post-processing techniques, we generated segmentation masks in a zero-shot framework. Our approach achieved an F1- score of 0.51, demonstrating the inherent challenges of spectrally complex coastal features. Despite this, the results highlight a promising pathway for adapting large-scale …

Llm-Assisted Legal Propositions Identification From Party Arguments In The U.S. Supreme Court Briefs, Heng Zheng, Alex Zhang

Faculty Scholarship

Merits briefs are central to U.S. litigation, serving as the primary means for parties to present arguments and persuade judges. Legal propositions in these merits briefs are the atomic units of arguments, whose relationships evolve throughout litigation and inform court decisions and precedent. Large language models (LLMs) have been applied to legal document review, but there is limited evidence on their ability to identify legal propositions in merits briefs. Given the labor-intensive nature of the task, we evaluate a human-AI collaborative approach to identifying legal propositions in the U.S. Supreme Court merits briefs, in which legal annotators review and revise …

Integrative Approaches And Data Analysis For Single-Cell Rna Sequencing Data, Teng Long

Computer Science and Engineering Dissertations

The rapid growth of single-cell RNA sequencing and transcriptomic datasets has created major computational challenges in causal discovery, representation learning, and biologically faithful data generation. To address these challenges, this dissertation presents three complementary deep learning frameworks for the analysis and modeling of transcriptomic data. Together, these methods form an integrative computational toolkit for understanding complex biological systems from high-dimensional and heterogeneous gene expression data.

First, this dissertation introduces DAG-VAERL, a causal discovery framework that integrates variational autoencoders, graph neural networks, reinforcement learning, and attention mechanisms to infer directed acyclic graphs for gene regulatory network analysis. DAG-VAERL improves causal structure …

Predicting Next-Day Eur/Usd Direction Using News Sentiment And Technical Indicators With Finbert And Xgboost, Darshan Sanjaybhai Khunt

Selected Full-Text Master Theses 2021-

Forecasting exchange-rate movements is a challenging task because currency prices are influenced not only by macroeconomic and financial variables but also by market sentiment reflected in financial news. This thesis examines whether financial news headlines can be used to predict the next-day directional movement of the EUR/USD exchange rate by applying finance-specific natural language processing and machine learning techniques.

The study uses a dataset of approximately 466,000 finance-related English-language news headlines collected between 2021 and 2025, aligned with daily EUR/USD closing prices. After preprocessing and temporal alignment, the data are used to construct a binary classification task in which the …

Reliable And Label-Efficient Learning For Open-World Visual Perception And Robot Learning Under Uncertainty, Zongyao Lyu

Computer Science and Engineering Dissertations

Modern learning systems deployed in open-world environments must make reliable decisions despite predictive uncertainty, previously unseen classes, limited annotations, and distribution shifts. This dissertation develops methods for reliable and label-efficient learning in visual perception and robot control.

First, this work studies uncertainty in object detection by representing semantic and spatial predictions probabilistically. A deep-ensemble framework aggregates detections into class-probability distributions and probabilistic bounding boxes, while a subsequent extension combines deep ensembles with Monte Carlo dropout to further investigate predictive uncertainty. Second, this dissertation addresses open-set recognition, where classes absent during training may appear at inference time. An empirical study shows …

Advancing Hate Speech Detection: Binary And Multiclass Approaches From Traditional Methods To Parameter-Efficient And Ontology-Guided Language Models, Mahmoud Abusaqer

Graduate Theses/Dissertations

The widespread proliferation of hate speech on social media platforms poses significant challenges for content moderation and user safety, requiring automated systems that are simultaneously accurate, efficient, and capable of fine-grained distinctions. This thesis investigates hate speech detection through five published manuscripts organized into two complementary threads: binary detection (hateful vs. non-hateful) and multiclass detection across demographic targeting categories. The binary thread progresses from a broad 38-model baseline spanning traditional machine learning, deep learning, and transformer architectures (where RoBERTa reaches 91.48% accuracy and CatBoost remains competitive at 88.60%) to parameter-efficient adaptation, in which Low-Rank Adaptation (LoRA) of large language models …