Physical Sciences and Mathematics Commons^™

Qlorax: Heuristic-Guided Fine-Tuning Of Llama-2 For Domain Adaptation In Entrepreneurship, Gaurob Saha

Theses and Dissertations

This thesis presents a study on the fine-tuning of large language models (LLMs) for domain-specific applications using limited data. We fine-tuned the LLaMA-2 (7B) model on a curated entrepreneurial dataset containing 3,545 human-written question-answer pairs, of which 3,095 were used for training and 450 were reserved for evaluation. A complete fine-tuning and evaluation pipeline was developed, which included clustering human-written answers, generating centroid-based summaries for each cluster, and evaluating the model's generated responses through cosine similarity.

Training was carried out over five epochs, with model performance evaluated after each epoch. The fine-tuned model demonstrated strong semantic alignment with human-written content, …

Exploratory Data Analysis (Eda) And Predictive Machine Learning (Ml) For Buildings’ Energy Fault Detection, Akshith Nukala

Theses and Dissertations

Building energy load fault detection is a critical challenge in energy usage analysis. It helps uncover energy wastage, machinery/appliance degradation or inefficiency, and failures or faults in buildings’ HVAC (heating, ventilation, and air conditioning) systems. Early identification of machinery failure and energy wastages due to operational maintenance negligence in large sites such as campus buildings is indispensable for achieving energy efficiency. This is crucial for saving patrol and minimizing the response time to restore the building appliances or systems to their optimal state.

Advancements in state-of-the-art AI/ML data-driven algorithms and techniques enabled us to build accurate, efficient and scalable fault …

Exploring Brent-Kung Adders In Balanced Ternary Cmos Logic, Astrea J.C. Rojas

Theses and Dissertations

The modern computer operates on a 64-bit architecture. These devices can store large numbers and precise decimals, but more advanced devices are needed to support progressing technologies every day. A more efficient system with higher speeds and larger operable numbers would be a key to optimization of computation as we know it. The ternary device, operating in base-3, has the potential to be that optimization. However, binary technology has such precedent and research that it is a difficult gap to span to compare the ternary system to the modern binary system. With a more advanced adder and optimized gates using …

An Analysis Of Coal Ash Pollution In The State Of Georgia, Mick W. Etchison

Theses and Dissertations

This social science paper will explore the factors contributing to the prevalence of coal ash in specific regions of Georgia and its impact on local ecosystems and human health. It provides an analysis of the current environmental and political issues surrounding the containment and spread of coal ash throughout the State. Coal ash is primarily found in the central, southwestern, and northeastern regions where coal-fired power plants and coal ash storage sites are located. The paper will then discuss the political, economic, and legal reasons behind the lack of effective regulation and mitigation of coal ash spread. Coal ash illustrates …

Assessing Equity, Environmental Social Justice And Intersectionality In U.S. Federal Policies: A Comparative And Temporal Analysis Of The Department Of Transportation, Department Of Energy, And Environmental Protection Agency, Sneha Basu

Dissertations and Theses @ UNI

As the climate crisis intensifies, U.S. federal policies play a critical role in addressing the disproportionate environmental burdens borne by marginalized communities. This thesis assesses how equity, environmental social justice, and intersectionality have been integrated into the policy documents of three key federal agencies: Department of Transportation (DOT), Department of Energy (DOE), and Environmental Protection Agency (EPA), over a 20-year period (2004–2024). Using a quantitative content analysis of 46 policy documents, the study evaluates five core dimensions: depth of discussion, clarity of language, inclusiveness of marginalized groups, action-oriented language, and environmental justice considerations. The research reveals significant interagency disparities, with …

Boy Whose Head Was Filled With Stars: A Life Of Edwin Hubble (2021) By Isabelle Marinov, Maughn Rollins Gregory

Picture Books

Barely a hundred years ago, Edwin Hubble changed the way we understand the universe and our place in it more dramatically than almost anyone else has ever done. As this picture book biography shows, his mind-boggling discoveries can be traced back to childhood questions. The discovery that the world is bigger than we had imagined happens in many ways, and learning that we have some choices in deciding how new facts, new experiences, and new perspectives can matter to us is a necessary part of education.

Online Learning-Driven Human Intent Estimation And Control For Human-Robot Interaction, Irfan Ganie, S. Jagannathan

Electrical and Computer Engineering Faculty Research & Creative Works

This paper presents a novel Stackelberg-game theoretic multilayer-online learning framework for cooperative control of nonlinear Physical Human-Robot Interaction (pHRI), where the human is modeled as the leader guiding a robot follower. This hierarchical interaction is captured as a dynamic Stackelberg game, with the human's intention estimated in real-time through online multilayer neural networks (MNNs). We introduce SVD-based weight update laws for actor-critic MNNs, which approximate value functions and control inputs for both human and robot, eliminating the need for predefined basis functions. In this framework, the human objective is first inferred and used to guide the robot actions by shaping …

Multi-Model Safe Neuro-Optimal Output Tracking Control Of Autonomous Surface Vessels With Explainable Ai, Behzad Farzanegan, S. Jagannathan

Electrical and Computer Engineering Faculty Research & Creative Works

This paper presents a safety-aware deep reinforcement learning (DRL)-based trajectory tracking control of autonomous surface vessels (ASVs). A multilayer neural network (MNN) observer estimates the ASV's state and uncertain dynamics. By utilizing the estimate state vector from the observer, a safety-aware DRL-based optimal policy is formulated using control barrier function (CBF) and Karush-Kuhn-Tucker (KKT) conditions. An actor-critic MNN with singular value decomposition (SVD)-based update mitigates vanishing gradients. To enhance adaptability, an online safe lifelong learning (SLL) scheme counters catastrophic forgetting across varying ASV dynamics. The Shapley Additive Explanations (SHAP) method identifies key features influencing the control policy. Simulations on an …

Online Adaptive Optimal Tracking Control Of Uncertain Strict Feedback Discrete-Time Systems With Hardware Verification Using A Quadrotor Uav, Maxwell Geiger, S. Jagannathan

Electrical and Computer Engineering Faculty Research & Creative Works

This article considers the infinite time horizon optimal tracking control problem for discrete time (DT) partially uncertain strict feedback systems with application to quadrotor UAVs. First, the strict feedback DT system is transformed into an equivalent affine nonlinear DT system in terms of tracking error dynamics. The optimal tracking control problem is solved using an augmented system approach, where a horizon of future reference trajectory points are used in the augmented state, as compared to using a single point. The internal dynamics of the original nonlinear strict feedback system and the transformed affine system in terms of error dynamics are …

Reinforcement Learning-Based Nonlinear Optimal Discrete-Time Control Of Power Systems, Vijay Kumar Singh, Behzad Farzanegan, S. Jagannathan

Electrical and Computer Engineering Faculty Research & Creative Works

This paper presents a partially model-free adaptive optimal tracking control method for power systems, specifically targeting a synchronous generator connected through a reactive transmission line. By integrating the tracking error dynamics with reference trajectory dynamics, an augmented system is created. A discounted performance function is introduced to address the nonlinear tracking problem optimally. Unlike traditional methods that compute feedforward and feedback terms separately, the proposed approach calculates both simultaneously by minimizing the discounted performance function. The discrete-time tracking Bellman and Hamilton-Jacobi-Bellman (HJB) equations are derived, and a reinforcement learning (RL)-based technique is employed to solve the optimal policy online without …

Ethics Vs.. Regulation: Converging Frameworks For Trustworthy Human-Centered Ai In Biomedical Research, Tayo Obafemi-Ajayi, Tiffani J. Bright, Emily F. Wong, Donald Wunsch, Joan Peckham, Jason H. Moore

Electrical and Computer Engineering Faculty Research & Creative Works

The accelerating impact of AI in biomedical research is driving significant advances in precision medicine. As these systems increasingly shape health outcomes, the imperative to develop trustworthy, reliable, and ethically grounded AI becomes more pressing, particularly in addressing concerns related to data integrity, patient safety, and equitable outcomes. While the potential of AI to transform biomedical research is clear, its responsible integration depends on more than technological capability. Ensuring that these systems are aligned with societal values requires a dual commitment: the operationalization of ethical principles throughout the AI life cycle and the establishment of robust regulatory mechanisms. Ethics provides …

Sift Feature-Based Relative Altitude Estimation Enhanced With Siamese Network, Shirin Nasr-Esfahani, S. Jagannathan

Electrical and Computer Engineering Faculty Research & Creative Works

In GPS-denied environments or when GPS signals are unreliable or unavailable, alternative methods of accurate localization with coordinate generation become critical. To address localization, the scale-invariant feature transform (SIFT) algorithm, along with its numerous adaptations, is extensively utilized in computer vision and remote sensing for matching image features to identify objects and perform localization. This article presents a novel approach for estimating the relative altitude of unmanned aerial vehicles (UAVs) using SIFT features' scale (size), omitting the need for additional data like camera intrinsic parameters, as well as extensive image datasets are also required for training. Furthermore, the approach enhances …

Foundation Models Boost Low-Level Perceptual Similarity Metrics, Abhijay Ghildyal, Nabajeet Barman, Saman Zadtootaghaj

Computer Science Faculty Publications and Presentations

For full-reference image quality assessment (FR-IQA) using deep-learning approaches, the perceptual similarity score between a distorted image and a reference image is typically computed as a distance measure between features extracted from a pretrained CNN or more recently, a Transformer network. Often, these intermediate features require further fine-tuning or processing with additional neural network layers to align the final similarity scores with human judgments. So far, most IQA models based on foundation models have primarily relied on the final layer or the embedding for the quality score estimation. In contrast, this work explores the potential of utilizing the intermediate features …

“Lost-In-The-Later”: Framework For Quantifying Contextual Grounding In Large Language Models, Yufei Tao, Adam Hiatt, Rahul Seetharaman, Ameeta Agrawal

Computer Science Faculty Publications and Presentations

Large language models (LLMs) are capable of leveraging both contextual and parametric knowledge but how they prioritize and integrate these sources remains underexplored. We introduce CoPE, a novel framework for systematically quantifying contextual grounding in LLMs. CoPE distinguishes between contextual knowledge (CK) and parametric knowledge (PK), enabling fine-grained attribution across languages and tasks. Using our newly created MultiWikiAtomic dataset in English, Spanish, and Danish, we analyze how LLMs integrate context, prioritize information, and incorporate PK in open-ended question answering. We find that across models and languages, only around 50 to 76 percent of outputs are grounded in the given context, …

Drta: Dynamic Reward Scaling For Reinforcement Learning In Time Series Anomaly Detection, Bahareh Golchin, Banafsheh Rekabdar, Kunpeng Liu

Computer Science Faculty Publications and Presentations

Anomaly detection in time series data is important for applications in finance, healthcare, sensor networks, and industrial monitoring. Traditional methods usually struggle with limited labeled data, high false-positive rates, and difficulty generalizing to novel anomaly types. To overcome these challenges, we propose a reinforcement learning-based framework that integrates dynamic reward shaping, Variational Autoencoder (VAE), and active learning, called DRTA. Our method uses an adaptive reward mechanism that balances exploration and exploitation by dynamically scaling the effect of VAE-based reconstruction error and classification rewards. This approach enables the agent to detect anomalies effectively in low-label systems while maintaining high precision and …

Scitopic: Enhancing Topic Discovery In Scientific Literature Through Advanced Llm, Pengjiang Li, Zaitian Wang, Xinhao Zhang, Ran Zhang, Lu Jiang, Pengfei Wang, Yuanchun Zhou

Computer Science Faculty Publications and Presentations

Topic discovery in scientific literature provides valuable insights for researchers to identify emerging trends and explore new avenues for investigation, facilitating easier scientific information retrieval. Many machine learning methods, particularly deep embedding techniques, have been applied to discover research topics. However, most existing topic discovery methods rely on word embedding to capture the semantics and lack a comprehensive understanding of scientific publications, struggling with complex, high-dimensional text relationships. Inspired by the exceptional comprehension of textual information by large language models (LLMs), we propose an advanced topic discovery method enhanced by LLMs to improve scientific topic identification, namely SciTopic. Specifically, we …

El Maíz Del Abuelo: Fidel Márquez Sanchez’S Legacy And Mexico’S Historical Maize Agriculture, Lucia Marquez-Uppman

Scripps Senior Theses

The Green Revolution was an era of scientific development in productive agricultural methods and technologies in the mid-twentieth century. High-yielding seed varieties, artificial fertilizers, irrigation infrastructure, pesticides, and herbicides all developed from this period. Texcoco, Mexico was the center of much of the Green Revolution's education and research from the 1940s onward at the Autonomous University of Chapingo, a prestigious agricultural school, with funding from the US-based philanthropic Rockefeller Foundation. Powerful consequences in the globe's agriculture emerged from the Green Revolution, including the production of maize, a staple crop in Mexico, especially for Indigenous and campesino populations. My abuelo, Fidel …

Random Graph Models For Dual Graphs, Anne Friedman

Scripps Senior Theses

This paper aims to better characterize dual graphs derived from state districting maps by developing random graph models that replicate their structural properties. Dual graphs provide a simplified way to represent districting maps, making it computationally feasible to analyze their structure. These representations enable researchers, legislators, and courts to assess district compactness, detect signs of gerrymandering, and generate alternative districting plans. A deeper understanding of the structural patterns of these dual graphs can help researchers choose or design more effective algorithms for redistricting analysis. The random graph models developed in this study serve as testbeds for evaluating algorithmic approaches to …

Weathering And Beyond: Leveraging Mathematical Modeling To Simulate Erosion In Digital Media, Fiona Irving-Beck

Scripps Senior Theses

How might we bring an idea to life from both a mathematical and an artistic perspective? Within Weathering, I use imagery of environmental erosion to explore the differences between physical and digital forms of representation. I created a physical painting of an abandoned copper mine, digitized the work, and then used a mathematical model to digitally “erode” it, which I re-translated into paintings. While the explicit texture present in physical work speaks best to my practice/intent, the mathematical framework that is the basis for my digital work affords a powerful mode of temporal flexibility. Used in conjunction, these two …

A Proof Of Np-Completeness For The K-Means Clustering Algorithm, Brooke C. Feinberg

Scripps Senior Theses

The k-means clustering algorithm is one of the most widely used clustering techniques in data analysis and machine learning, yet its exact computational complexity remains subject to ongoing theoretical investiga- tion. This work establishes the NP-completeness of k-means by proving (1) it is NP-hard and (2) it lies in NP. To demonstrate NP-hardness, we construct a series of polynomial-time reductions from well-known NP-complete problems. Specifically, we reduce 3sat to Vertex Cover, and then reduce Vertex Cover to k-means, thereby establishing the computational hardness of the k-means clustering problem. We then prove k-means is in NP, and thus conclude it is …

Empirical Analysis Of Political Districting Splitability Via Uniform Spanning Trees In Polynomial Time, Brooke C. Feinberg

Scripps Senior Theses

This work expands a recently proven conjecture that a polynomial fraction of all uniform spanning trees (USTs) are splittable into k balanced partitions on grid graphs to real-world political districting plans. We investigate whether similar structural properties hold for the planar dual graphs of U.S. counties (cnty) and tracts (t), using Wilson’s algorithm to generate uniform random spanning trees and Breadth- First Search (BFS) to check for splitability into balanced partitions. Our empirical findings suggest that real-world districting plans can be split into 2-balanced, connected partitions in a fraction of polynomial time. This result highlights the potential for scalable redistricting …

Explainable And Safety Aware Deep Reinforcement Learning-Based Control Of Nonlinear Discrete-Time Systems Using Neural Network Gradient Decomposition, Behzad Farzanegan, S. Jagannathan

Electrical and Computer Engineering Faculty Research & Creative Works

This paper presents an explainable deep-reinforcement learning (DRL)-based safety-aware optimal adaptive tracking (SOAT) scheme for a class of nonlinear discrete-time (DT) affine systems subject to state inequality constraints. The DRL-based SOAT utilizes a multilayer neural network (MNN)-based actor-critic to estimate the cost function and optimal policy while the MNN update laws are tuned both using the singular value decomposition (SVD) of activation function gradient in order to mitigate the vanishing gradient issue and safety-aware Bellman error at each layer. An approximate safety-aware optimal policy is developed using Karush–Kuhn–Tucker (KKT) conditions by incorporating the higher-order control barrier function (HOCBF) into the …

Training Neural Networks With A Self-Adaptive Ant Colony Algorithm, Ashraf M. Abdelbar, Donald C. Wunsch

Electrical and Computer Engineering Faculty Research & Creative Works

ACOR is a well-established ant colony optimization algorithm that has been applied to neural network training. We present an approach for the dynamic adaptation of the ACOR algorithm's search intensification/diversification parameter q, based on using several pre-specified parameter configurations, which we call personalities. Before an ant begins to generate a candidate solution, it stochastically adopts a personality based on the relative past success of the different personalities. The success of a personality is measured, in turn, by the relative quality of previous solutions generated by ants adopting that personality. The premise of our approach is that some personalities will be …

Utilizing Artificial Intelligence As A Strategic Risk Management Tool For Public Sector Operations And Auditing Processes, Oğuz Ümit Tamer, Bruce D. Mcdonald Iii, Farouk Hemici, Georgia Kontogeorga

School of Public Service Faculty Publications

Symbolizing a significant turning point in the historical landscape, AI is becoming an effective tool in today's public administration, not only for increasing capacity, quality, and speed in services, but also for strategic risk management. Regulators and algorithmic auditing play a central role in implementing fairness, transparency, and persistent controls against risks in AI systems. Discussing modern applications of AI, such as anomaly-based fraud detection, resource estimation, and continuous auditing, and their respective strengths and weaknesses, this study concludes that AI significantly enhances efficiency and oversight but also poses the risk of enshrining bias, opacity, and accountability gaps. By considering …

An Understanding Of The Feeding Ecology Of Teleoceras Fossiger And Teleoceras Aepysoma Using Dental Microwear And Mesowear Analysis, Gabriel S. Goncalves

Master's Theses or Doctor of Nursing Practice

Teleoceras is a genus of rhinocerotids that belongs to the Tribe Teleoceratines and was widespread across North America from the late Early Miocene into the Early Pliocene. This genus was very successful, spreading across the continental United States from Florida and Tennessee to California and Washington state down to central Mexico and Honduras. Teleoceras is characterized by having a large, rounded chest cavity and short limbs with relatively high crowned teeth and lower incisors modified into tusk-like teeth. Initial ecological comparisons were made to the modern hippopotamus based on proportions of the chest cavity and its hypsodont dentition. Stable carbon …

Geochemical Assessment Of Sediment Flux In A Sediment Core From A Tidal Flat In A Tropical Estuary, Olumide Ajulo

Master's Theses or Doctor of Nursing Practice

Sediment flux in estuaries is influenced by both natural processes and anthropogenic activities. Understanding the historical and current trends in sediment flux inputs and geochemical composition are essential for evaluating the estuary’s response to natural and human-induced changes and for developing effective environmental management strategies. The tropical Wouri Estuary in Cameroon receives water and sediment inputs from three main rivers draining urban, industrial, and agricultural areas, as well as different rock terrains. The complex inputs into the estuary can significantly affect sediment flux, consequently impacting the ecosystem. Thereby making an understanding of the temporal variations in sediment inputs in the …

Video Comprehension Score (Vcs): A Metric For Long-Form Video Description Evaluation, Harsh Dubey

Electronic Theses and Dissertations

Existing video description evaluation metrics fail to capture the long-range chronology and semantic alignment essential for long-form descriptions. An effective evaluation metric for long-form descriptions must (i) assess global thematic alignment, (ii) measure local semantic alignment, and (iii) evaluate chronological alignment while detecting corrupted content. We introduce Video Comprehension Score (VCS), a reference-based metric, which directly addresses these evaluation requirements through three components: Global Alignment Score for thematic alignment, Local Alignment Score for local semantic alignment, and Narrative Alignment Score for chronological alignment with adjustable tolerance. We evaluate VCS on two large-scale synthetic datasets designed to test corruption detection and …

Quantifying Landscape Variability Of Dynamic Soil Properties In The James River Valley, Eastern South Dakota, Sage Reuter-Schmitt

Electronic Theses and Dissertations

Dynamic soil properties (DSP) are a great tool for monitoring land-use change, and they are frequently used as soil health metrics. The NRCS and other soil researchers often use repeating, single-point or nested-point DSP studies to monitor and better understand the effects of land management decisions. While these studies capture the temporal variability of DSPs, they overlook the field-scale spatial variability of DSPs. Soil organic carbon (SOC) provides critical services to the ecosystem and other environments. These traits, as well as its susceptibility to management decisions, make SOC an important soil health indicator and DSP. Measuring SOC is often time …

The Spatial And Temporal Dimensions Of Plant–Soil Feedback In Native Perennial Monocultures, Bret Lang

Electronic Theses and Dissertations

Plant-soil feedbacks (PSFs) describe how plants change their soil environment in ways that can either negatively or positively affect subsequent plant growth. In doing so, they offer an opportunity to connect ecological processes with agricultural practice. My dissertation examines PSFs created by native perennial monocultures, commonly used for native seed production, across time and space, and through soil legacies. Specifically, I focus on how plant species identity shapes both soil nutrient and soil microbial community responses. To address these questions, I conducted field experiments with monocultures of five forb species native to South Dakota and established plots with these species …

Survey-Weighted Ordinal Modeling Of Alcohol-Associated Liver Disease Severity Through Social Determinants Of Health, Jaylene Viveros Cruz

Selected Full-Text Master Theses 2021-

Alcohol-associated liver disease (ALD) is a condition that describes the spectrum of disease and liver injury attributed to the consumption of alcohol. ALD diagnosis is heavily dependent on alcohol use, making it challenging to test for, as alcohol use is often self-reported, and early-stage ALD can present asymptomatically. This study aims to explore how social determinants of health associated with alcohol use behaviors predict ALD-related liver stress risk. MEC participants included in the two two-year cycles of the National Health and Nutrition Examination Survey (NHANES), 2014-2014 and 2015-2016, were assigned liver stress labels based on clinical thresholds for ALD diagnosis. …