Computer Sciences Commons^™

Closed Domain Question Answering With Language Models: Application Of Retrieval-Augmented Generation And Parameter Efficient Fine-Tuning In Healthcare, Aaron Cummings

Master's Theses

Dementia care presents significant challenges for informal caregivers, particularly in managing behavioral symptoms that affect over 90% of individuals with Alzheimer’s Disease and Related Dementias (ADRD) during the moderate-to-severe stages. These symptoms, including agitation, wandering, and repetitive activities, impose emotional and physical burdens on caregivers, often exacerbated by a lack of reliable, accessible, and personalized resources. Non-pharmacological interventions, while evidence-based, are underutilized due to knowledge gaps and the inefficiency of traditional training and information retrieval methods.

This research explores the adaptation of large language models (LLMs) to address these challenges by developing a framework for closed-domain Question Answering (QA) systems, …

Gpu Exploration Using Pycuda, Prudhvi Sai Akulapally

Masters Projects

This project investigates the implementation and optimization of parallel computing techniques using GPU acceleration frameworks such PyCuda. With the increasing demand for high-performance solutions in data-intensive applications, GPUs offer a compelling alternative to traditional CPU-based processing. The primary objective of this work is to harness the computational power of GPUs to achieve significant performance enhancements for complex workloads. The study focuses on two essential research questions: How effectively can PyCuda accelerate computational tasks, and what measurable performance gains can be achieved compared to CPU-based implementations? By addressing these questions, this project explores the applicability of GPU programming in tasks such …

Tag-Based Security In C: Writing And Specifying Flexible Protection, Sean Noble Anderson

Dissertations and Theses

The C language is ubiquitous and insecure. Tag-based security policies offer a flexible toolkit for runtime protection, including policies that mitigate the effects of undetected programming bugs and those that enforce security properties of the program logic. But tag policies can be difficult to define, and the protection that they offer can be difficult to specify and to prove.

This dissertation builds from an assembly-level specification of stack safety, encompassing the essential control-flow structure of C and other high level languages, to a C source-level policy definition framework called Tagged C. Tagged C includes a variant C semantics parameterized by …

A Survey Of Several Python Libraries Foe Computational Geometry, Meghana Kolluru

Masters Projects

This project delves into computational geometry, a crucial area of computer science, by

implementing algorithms to solve geometric problems using Python libraries like Shapely, SciPy,

and Triangle. It begins with foundational tasks such as creating geometric objects (points, lines,

polygons) and performing operations like calculating distances, intersections, and unions. These

basics lay the groundwork for tackling more advanced applications.

Key implementations include the Convex Hull, which computes the smallest convex

polygon enclosing a set of points, aiding in applications like collision detection and GIS. The

project also explores Voronoi Diagrams, which partition a plane into regions based on proximity,

and …

Page Rank Algorithm In Java, Nikkil Bollman

Masters Projects

The PageRank algorithm is introduced by Google, It is used in search engine optimization, and ranking web pages based on their significance and importance based on a graph structure. This project focuses on implementing the PageRank algorithm using Java and Spring Boot, with Apache Spark as the core technologies for graph data processing. The project aims to compute PageRank scores for nodes in a directed graph, simulating real-world scenarios such as web page ranking and link analysis. The implementation is designed to handle large datasets efficiently by leveraging Spark's distributed computing capabilities and GraphFrames' robust graph processing framework. The workflow …

Creating A Colab Notebook Based On The Tutorial Of The Simpy Python Library, Mary Deepika Basani

Masters Projects

This project introduces an interactive Google Colab notebook designed to familiarize users with the core functionalities of the SymPy Python library, a powerful tool for symbolic computation. SymPy offers an extensive range of capabilities essential for mathematical modeling, algebraic operations, calculus, and matrix manipulations. The notebook serves as an educational platform, allowing users to experiment with small examples and explore practical applications of symbolic computation in real time. Organized into sections, the notebook focuses on key SymPy features such as simplification, equation solving, differentiation, integration, and matrix operations. Each section includes clear explanations, practical demonstrations, and interactive exercises, helping users …

Real-Time Feedback-Driven Framework For Automated Cybersickness Mitigation, Md Jahirul Islam

Master's Theses

As technologies are becoming more advanced day by day, the embracement of virtual reality (VR) technology among users is also increasing in daily activities for various purposes, and subsequently, the barrier between the real and virtual world is fading. Despite the versatile uses, cybersickness (CS) is a major problem which is induced among users due to the immersive VR experience. There is a plethora of research findings and methods to measure the users’ CS such as virtual reality sickness questionnaire (VRSQ), simulator sickness questionnaire (SSQ), fast motion scale questionnaire (FMS), and others. Recently, machine learning approaches have also been adopted …

Exfoliated Hydrotalcite-Transition Metal Complex Composite For Eco-Friendly And Efficient Catalytic Degradation Of 4-Nitrophenol, Sidra Khan, Najma Memon, Saima Q. Memon, Yilmaz Yurekli

Karbala International Journal of Modern Science

Nitrophenols are notorious aquatic organic contaminants found as degradation products of various parent compounds, including pesticides and industrial chemicals that persist in the environment and must be removed. Catalytic degradation is one of the feasible routes to clean the contaminated water systems, however, environmental contamination with catalysts is also widespread. Herein, we report an environmentally friendly catalyst based on composited Fe-Schiff’s base with exfoliated layered double hydroxides (LDH) of aluminum and nickel (hydrotalcite). The composite showed agglomerated pleated LDH structures sheathed with Fe(III)SB. Nitrogen adsorption isotherm data exhibited improved surface area and narrow pores patterns for composite as compared to …

Optimal Algorithm For Managing On-Campus Student Transportation, Youssef Harrath, Jood Alyusuf, Zeena Ghulam, Muna Aldoseri

Research & Publications

This study analyzed the transportation issues at the University of Bahrain Sakhir campus, where a bus system with an unorganized and fixed number of buses allocated each semester was in place. Data was collected through a survey, onsite observations, and student schedules to estimate the number of buses needed. The study was limited to students who require to move between buildings for academic purposes and not those who choose to ride buses for other reasons. An algorithm was designed to calculate the optimal number of buses for each time slot, and for each day. This solution could improve transportation efficiency, …

Enhancing Cybersecurity Strategies Through Automated Cti Extraction, Risk Prioritization, And Privacy-Conscious Information Sharing, Spencer Rian Massengale

Theses and Dissertations

Cybersecurity operations require the ability to collect and analyze large amounts of cyber threat intelligence (CTI) to assess risks and formulate defensive strategies against emerging threats. This task has become increasingly complex due to the rapid evolution of cyber threats and the growing volume of unstructured, natural-language CTI sources. The scale of data and analysis needed to utilize CTI effectively far exceeds humans' manual capacity, especially for organizations with limited resources. This research focuses on leveraging Large Language Models (LLMs) and machine learning techniques to enhance CTI extraction, risk assessment, and data sharing. We utilized LLMs to automate the extraction …

How Does Augmentation Affect Feature Space: A Study Using Various Augmentation Methods In Distributed Learning, Nikil Sharan Prabahar Balasubramanian

Computer Science Theses

This thesis examines the impact of data augmentation techniques on model performance within a distributed learning framework, focusing on enhancing feature diversity and improving representation for under-represented classes. Data augmentation, commonly used to address data imbalance, significantly influences the feature space learned by deep learning models, with varied effects in distributed settings where data is split across nodes. Our study reveals that inconsistencies in feature learning across nodes reduce the benefits of local augmentation in capturing complex patterns, leading to suboptimal model performance. To address this, we propose a coherent augmentation approach that embeds consistent transformations in the central server, …

Predicting Chaotic Systems With Quantum Echo-State Networks, Erik Connerty, Ethan N. Evans, Gerasimos Angelatos, Vignesh Narayanan

Publications

Recent advancements in artificial neural networks have enabled impressive tasks on classical computers, but they demand significant computational resources. While quantum computing offers potential beyond classical systems, the advantages of quantum neural networks (QNNs) remain largely unexplored. In this work, we present and examine a quantum circuit (QC) that implements and aims to improve upon the classical echo-state network (ESN), a type of reservoir-based recurrent neural networks (RNNs), using quantum computers. Typically, ESNs consist of an extremely large reservoir that learns high-dimensional embeddings, enabling prediction of complex system trajectories. Quantum echo-state networks (QESNs) aim to reduce this need for prohibitively …

Identifying Subject Bias In Wifi-Based Human Activity Recognition Evaluation Methods, Amany Elkelany, Robert J. Ross, Susan Mckeever

Conference papers

WiFi-based Human Activity Recognition (HAR) has emerged as a promising approach for monitoring and analysing human activities in a non-intrusive manner, leveraging WiFi signals for activity classification. Despite advancements, existing WiFi-based HAR research lacks consideration of subject (human) bias. This results in learning models performing well on individuals used in the training samples but failing to generalise to new/unseen subjects, in contrast to known good practices in machine learning. In this paper, we address this oversight directly by systematically examining the evaluation methodology for the WiFi-based HAR context. Specifically, we investigate the impact of Leave-One-Subject-Out Cross-Validation (LOSOCV) in a hybrid …

Sd-Weat: Towards Robustly Measuring Bias In Input Embeddings For Artificial Intelligence Language Models, Magnus Gray

Theses and Dissertations

Artificial intelligence (AI) is rapidly transforming industries and markets, from healthcare to entertainment, revolutionizing decision-making processes. However, as AI grow more influential, they also risk amplifying existing biases, potentially leading to harmful consequences. Recent advancements in large language models (LLMs), such as GPT-4 and Llama, have heightened concerns about bias in natural language processing (NLP) tasks, driving the need for robust methods to detect and mitigate bias. Current approaches, such as the Word Embedding Association Test (WEAT) and its sentence-level extension the Sentence Encoder Association Test (SEAT) often fall short in capturing the nuances of biases in the input embeddings …

On The Benefits Of Directness In Virtual Characters For Motivational Interviews, Michael O'Mahony, Cathy Ennis, Robert Ross

Conference papers

Understanding the factors influencing successful engagement with Embodied Conversational Agents (ECAs) remains a significant challenge. This understanding could be used to personalise agents to users to improve interactions. Some studies have shown that simulating personalities in healthcare agents can improve effectiveness and engagement. However, it is not yet well understood how variations of agent personality can be leveraged to improve user engagement with Motivational Interviewing (MI) ECAs. Specifically how the balance between agent warmth and directness can be controlled in an MI agent to improve likeability and engagement. We conducted an online Wizard-of-Oz (WoZ) mediated study of two variants of …

A Summer Class Exploring Computer Science With Educators Of Deaf And Hard Of Hearing Students, Meghan L. Mcsherry, Becca A. Leininger, Maria L. Johnson, Annmarie P. Thomas, Susan Outlaw, Douglas C. Orzolek

Journal of Science Education for Students with Disabilities

The purpose of this research project was to design, deliver, and informally assess the content and methodology of an introductory course focused on the block-based coding language, Scratch, for educators of Deaf and Hard of Hearing students (N=20). Prior work by The Playful Learning Lab examined various STEM resources and their effect on student perception of STEM topics and content retention when utilized for Deaf and Hard of Hearing K-12 students. Previous research has shown that computer science and programming resources and curricula available today are not fully accessible for Deaf and Hard of Hearing K-12 students. Educators and teachers …

Unpacking Bias, Accountability, And Ethical Practices In Ai, Manya Chandra, Micol Hebron

Student Scholar Symposium Abstracts and Posters

This study is based on understanding how text-to-image generative AI platforms perpetuate biases such as racism and sexism and decoding how this bias is programmed within large language models and datasets. In this study, the results of generative AI are analyzed through the lens of affect and affect theory, as they are applied to investigate the machine learning and computer theory behind generative AI algorithms. The purpose of the study is to explain why generative AI is biased and whether this bias is generated due to current trends or to deficits and biases within the database that it draws information …

Dancetag: Using Sensors To Improve Feedback Given To Dance Students, Yanelly Mego, Franceli L. Cibrian

Student Scholar Symposium Abstracts and Posters

The structure of dance classrooms has remained largely unchanged for years, with minimal integration of technology to enhance teaching. This has motivated our research project, which aims to capture dance movements using wearable sensors and translate the information into meaningful visualizations to help dancers improve their skills. As the first step in addressing the research question—can data from commercial wearables differentiate between the movements of dancers and non-dancers?—we developed DANCETAG (Data Analytics and Notation with Captured Event Tagging), a platform designed for data collection and movement annotation. We utilized Sony’s Mocopi sensors, a motion capture system with six sensors attached …

Design Implications For Montessori Education, Olivia Chilvers

Student Scholar Symposium Abstracts and Posters

This project aims to identify the design implications for children’s educational apps that seek to align with the principles of the Montessori Philosophy and contribute to the development of educational technology. By analyzing critical features of apps that engage children and promote their natural desire to learn, this project aligns with the Montessori philosophy of fostering independent, self-directed learning. To uncover the design implication, we conducted a qualitative analysis of reviews of popular educational apps, defined by having at least 3,000 reviews and a rating of 3 stars or higher, that appeared from search results using the keyword “Montessori” in …

Software Implementations And Analyses Of The Emotional Impact Of Various Binaural Beat Classifications Layered Into Music., Neil Azimi

Student Scholar Symposium Abstracts and Posters

This study explores the psychoacoustic effects of binaural beats, which are produced when sinusoidal waves of slightly differing frequencies are played into each ear, leading to brainwave entrainment. Binaural beats are categorized by frequency bands (e.g., Beta: 14–30 Hz for energy, Theta: 4–8 Hz for relaxation), each associated with different psychological effects. This research contributes to the field by empirically analyzing whether binaural beats alter the emotional impact of music. Past studies have investigated the potential benefits of binaural beats in relaxation and energy stimulation. However, their effects, when combined with music, especially regarding a song's perceived emotional quality or …

Lstm-Transformer Based Robust Hybrid Deep Learning Model For Financial Time Series Forecasting, Md R. Kabir

Theses and Dissertations

The inherent challenges of financial time series forecasting demand advanced modeling techniques for reliable predictions. Financial data, belonging to the category of multimedia data, contains an extensive quantity of information that is widely used for data analysis and decision-making tasks. Effective financial time series forecasting is crucial for financial risk management and the formulation of investment decisions. Accurate prediction of stock prices is a subject of study in the domains of investing and national policy. This problem appears to be challenging due to the presence of multi-noise, nonlinearity, volatility, and chaotic natures in stocks. This paper proposes a novel financial …

Developing A Nearly Automated Open-Source Pipeline For Conducting Computational Fluid Dynamics Simulations In Anterior Brain Vasculature: A Feasibility Study, Mostafa Rezaeitaleshmahalleh, Nan Mu, Zonghan Lyu, Joseph Gemmete, Aditya Pandey, Jingfeng Jiang

Michigan Tech Publications

Intracranial aneurysms (IA) pose significant health risks and are often challenging to manage. Computational fluid dynamics (CFD) simulation has emerged as a powerful tool for understanding lesion-specific hemodynamics in and around IAs, aiding in the clinical management of patients with an IA. However, the current workflow of CFD simulations is time-consuming, complex, and labor-intensive and, thus, does not fit the clinical environment. To address these challenges, we have developed a semi-automated pipeline integrating multiple open-source software packages to streamline the CFD simulation process. Specifically, the study utilized medical angiography data from 18 patients. An in-house open-source DL image segmentation model …

Applying Positive Unlabeled Learning Techniques And Using The Kullback-Leibler Divergence To Improve Geothermal Surveying Assessments, Martín Thomas Rodriguez

Dissertations and Theses

As we face the current climate crisis, the discovery of geothermal energy resources has the potential to greatly reduce our dependence on fossil fuels worldwide. However, the development of any new energy infrastructure is expensive and depends on the willingness of energy agencies and developers to make initial investments based on calculated risk measures. One such measure, called geothermal favorability, is the likelihood that a site has conditions favorable for geothermal systems containing recoverable energy potential. Its prediction from existing geophysical datasets proves to be a nontrivial task. The prediction of geothermal favorability can be framed as a binary classification …

The Computational Eye. Deconstructing Style In Digital Art History, Paul Guhennec, Ellen Charlesworth

Artl@s Bulletin

With the aim of grounding digital methods in the art historic tradition, this paper uses the discussions around style as a springboard to ask how digital art history can extend beyond providing quantitative confirmation of known trends to enrich our current understanding of visual cultures. Drawing from the examples throughout this issue, we explore how an analysis of computational ways of seeing—or the ‘computational eye’—can expose the underlying preoccupations and priorities of our own research.

Afin de mieux ancrer les méthodes numériques dans la tradition de l’histoire de l’art, cet article se sert des discussions récentes autour du concept de …

Visualization Of Paleocurrents On A Web Application Using Gplates, Anjan Sapkota

MS in Computer Science Theses

Paleocurrents are flow directions derived from features of sedimentary rocks that reveal the direction of the current of wind or water that deposited the sediment. In 2015, Brand et al. created a global database of paleocurrents, which contains over 1,000,000 measurements worldwide: North America, South America, Australia, Great Britain, parts of Western Europe, China, Africa are fairly well represented; Antarctica, Eastern Europe, and Asia are modestly represented and Russia is poorly represented. The contribution of this thesis is a web application that uses the GPlates’ Application Programming Interface (API) to visualize global paleocurrents through time in an interactive way based …

Artificial Creativity: How Artificial Intelligence Will Impact Creativity Via Post-Production, Nicole Dwyer

Honors Thesis

My thesis lives in the world of Post-Production, and it contains both a creative and written component. I was the editor for four Undergraduate thesis projects. Seeking to gain experience in editing various genres, I worked in drama, sports, period piece, coming of age, and adventure. My biggest takeaways from these projects are the importance of organization, communication, time management, and editing with a sense of imagination. I became a more confident, resilient, and prepared editor through these experiences.

Along with the hands-on filmmaking element of my thesis, I also conducted research on how artificial intelligence will impact conceptions of …

Graph Neural Networks Powered Scientific Paper Recommendation, Junhao Shen

Computer Science and Engineering Theses and Dissertations

Scientific paper recommendation systems aim to help researchers discover relevant papers amidst the vast and ever-growing body of literature. With the exponential yearly increase in scientific publications, the demand for effective paper recommendation solutions has become both critical and increasingly challenging. In recent years, deep learning techniques have revolutionized recommender systems, and scientific paper recommendations have naturally integrated these advancements. In this dissertation, we address these challenges through three progressive contributions.

First, we enhance traditional content-based methods using Graph Neural Networks (GNNs) by introducing a Graph Convolutional Network-strengthened Topic Modeling (GCN-TM) approach. This method improves upon conventional topic modeling techniques …

Machine Visual Perception From Sim-To-Real Transfer Learning For Autonomous Docking Maneuvers, Derek Worth, Jeffrey Choate, Ryan M. Raettig, Scott L. Nykl, Clark N. Taylor

Faculty Publications

This paper presents a comprehensive approach to enhancing autonomous docking maneuvers through machine visual perception and sim-to-real transfer learning. By leveraging relative vectoring techniques, we aim to replicate the human ability to execute precise docking operations. Our study focuses on autonomous aerial refueling as a use case, demonstrating significant advancements in relative navigation and object detection. We introduce a novel method for aligning digital twins using fiducial targets and motion capture data, which facilitates accurate pose estimation from real-world imagery. Additionally, we develop cost-efficient annotation automation techniques for generating high-quality You Only Look Once training data. Experimental results indicate that …

Website Portfolio, Thomas Roch

Honors Program Theses and Projects

No abstract provided.

Alphafold2-Based Characterization Of Apo And Holo Protein Structures And Conformational Ensembles Using Randomized Alanine Sequence Scanning Adaptation: Capturing Shared Signature Dynamics And Ligand-Induced Conformational Changes, Nishank Raisinghani, Vedant Parikh, Brandon Foley, Gennady M. Verkhivker

Mathematics, Physics, and Computer Science Faculty Articles and Research

Proteins often exist in multiple conformational states, influenced by the binding of ligands or substrates. The study of these states, particularly the apo (unbound) and holo (ligand-bound) forms, is crucial for understanding protein function, dynamics, and interactions. In the current study, we use AlphaFold2, which combines randomized alanine sequence masking with shallow multiple sequence alignment subsampling to expand the conformational diversity of the predicted structural ensembles and capture conformational changes between apo and holo protein forms. Using several well-established datasets of structurally diverse apo-holo protein pairs, the proposed approach enables robust predictions of apo and holo structures and conformational ensembles, …