Physical Sciences and Mathematics Commons^™

Building Trustworthy Information Systems: A Unified Framework For Comparative Risk Detection, Parisa Momeni

USF Tampa Graduate Theses and Dissertations

Risk detection in large scale information systems increasingly depends on heterogeneous data generatedby both centralized and distributed ecosystems. While centralized systems provide curated and validated reports, distributed environments produce large-scale and real-time observational evidence. Existing computational approaches analyze these ecosystems in isolation, limiting systematic comparison of risk repre-sentations across heterogeneous sources.

This dissertation presents a unified computational framework for comparative risk detection across centralized and distributed information systems. The framework provides a domain independent methodology for transforming heterogeneous risk reporting data into comparable multidimensional representations. To enable interpretable comparison of heterogeneous risk distributions, this work introduces the Geometric Overlap Score …

Seeing Around Corners: An Indirect Light Transport Decomposition Framework For Fusing Physics And Learned Priors, Fadlullah Raji

USF Tampa Graduate Theses and Dissertations

Non-line-of-sight (NLOS) imaging, or seeing around corners, is the ability to recover information about objects hidden from direct view and remains one of the most compelling challenges in computational imaging. Existing approaches fall broadly into two categories: active and passive. Active time-resolved methods achieve impressive performance but require specialized pulsed laser hardware and ultrafast detectors, making them expensive and limited by slow measurement acquisition. Passive NLOS methods, which instead exploit the subtle soft shadows (penumbrae) cast by a hidden scene onto a visible matte surface, have recently emerged as a practical alternative. However, existing passive approaches are largely limited to …

Language Models For Oncology Clinical Text: How Model Architecture And Data Strategies Shape Tumor Phenotype Extraction And Disease Progression Detection, Thanh Duong

USF Tampa Graduate Theses and Dissertations

The rapid growth of electronic health records (EHRs) has created new opportunities to apply machine learning to clinical data.However, a large portion of important clinical information is still stored in unstructured text, such as pathology reports, radiology reports, and longitudinal clinical notes.These documents contain key details about tumor characteristics, diagnoses, treatments, and patient outcomes.Extracting structured and useful information from this text is challenging due to complex medical language, varied document formats, and the need to combine information across multiple reports over time.This dissertation studies how language models can be designed and adapted to better extract and use oncology-specific information from …

Mapping Seafloor Habitats And Characterizing Demersal Reef Fish Communities On The West Florida Shelf Across Multiple Spatial Scales Using Multibeam Sonar And Underwater Video, Alexander R. Ilich

USF Tampa Graduate Theses and Dissertations

In this dissertation, I examine the role of spatial scale in deriving seafloor terrain attributes, modeling benthic habitat, and analyzing ecological patterns of the demersal reef fish communities on the West Florida Shelf (WFS), with implications for improving predictive habitat mapping and understanding fish-environment relationships. Spatial scale plays a central role in benthic habitat mapping and ecological inference, influencing how terrain attributes are derived, how habitats are modeled, and how ecological responses are manifested. This work integrates multiscale terrain analysis, acoustic remote sensing, and community ecology to evaluate how spatial scale influences both physical habitat representation and ecological analysis.

Chapter …

A Mathematical Model Of The Interactions Between A Collagen Lattice, Fibroblasts, And Fixed Surfaces With Varying Topographies, Mary Jenkins

Theses and Dissertations

Collagen is an important structural protein in the body, which plays a role in wound healing, particularly the contraction process. Collagen lattices have been studied for nearly 50 years to provide insight into wound contraction. In some cases, collagen interacts with surfaces that have fixed shapes, such as medical implants. Our model focuses on the interactions of a collagen lattice with such a fixed surface. We mathematically model a collagen lattice as a network of nodes connected by springs. We also model fixed surfaces that have various topographies. Our model includes fibroblast cells that connect to both surfaces and remodel …

An Analysis Of Heat Spread Through Smoldering Pine Needles, Timothy Keith

Theses and Dissertations

The goal of this thesis is to calculate the average speed of flame spread through smoldering pine needles. We present two different models for doing this: the first is a very large system of ODEs meant to represent the spread of flame through individual needles, the second is a macroscopic model based on the porous medium PDE. Simulations are run for each of these models and compared to each other as well as to existing models and experimental results.

Portable Trace Gas Analysis With Multipass Cavity Raman Scattering, Charuka Muktha Arachchige

USF Tampa Graduate Theses and Dissertations

The continual increase in the production and utilization of chemicals across a growing range of industrial, environmental, and medical applications necessitates accurate and efficient measurement of molecular composition. Trace gas detection is particularly critical in industrialized settings, where the identification of colorless and odorless gases is essential for ensuring operational safety and regulatory compliance. Recent advances in optical instrumentation have positioned spontaneous Raman scattering as a promising analytical technique for trace gas analysis, with multipass cavity enhancement providing a cost-effective pathway to high detection sensitivity. However, existing implementations remain confined to laboratory-scale optical table setups, limiting their applicability in field …

Designing For Fitness And Resilience In Human Artificial Intelligence Systems, Thomas R. Gill

USF Tampa Graduate Theses and Dissertations

Over the last decade, new artificial intelligence models, frameworks, and applications have emerged that have radically changed the socio-technical landscape. These new technologies represent both great opportunities and great potential risks. Due to the recency of their emergence, there is little known about the longevity and long-term repercussions of artifacts powered by new artificial intelligence technology. In this dissertation, I evaluate A.I. and A.I.-based artifacts through the lens of the fitness-utility model, a tool developed by Gill and Hevner (2013) for the evaluation of an artifact's ability to survive and reproduce over long time horizons. In so doing, I also …

Non-Euclidean Geometries And Fairness Constraints In Advanced Clustering, Arnab Seal

Master’s Dissertations

A fundamental challenge in modern unsupervised learning is adapting classical clustering algorithms to handle complex, real-world data constraints. Traditional models often assume data resides in a flat, Euclidean space and optimize strictly for cluster cohesion, thereby failing to capture intrinsic hierarchical structures and ignoring sociotechnical demographic biases. This thesis addresses these critical limitations by extending generalized mean-shift dynamics into two novel clustering frameworks. First, to natively accommodate data with tree-like structures (e.g., taxonomies and social networks), we propose Hyperbolic Gaussian Blurring Mean Shift (HypeGBMS). By projecting data into the Poincar´e ball model and utilizing M¨obius vector space operations, HypeGBMS successfully …

Leveraging Spatial Statistics For Domain Adaptation Of Vision Language Models In Medical Vqa, Himanshu Raj

Master’s Dissertations

Recent advances in Vision–Language Models (VLMs) have demonstrated strong performance in Medical Visual Question Answering (Medical VQA) task. Although they perform very well within their domains, these models often experience issues with their generalization ability on unknown clinical distribution data because of different imaging technologies and patient groups used in various medical facilities. Generalization problems faced by these models make their practical application in the field of VLM-based medical VQA systems rather difficult. To overcome this limitation we proposed our method named Spatial Semantics Aware Domain Adaptation (SSADA), which is an integrated framework that combines both finetuning and prompt-based in-context …

Predictive Importance Sampling Based Coverage Verification For Multi Uav Trajectory Planning, Snehashish Ghosh

Master’s Dissertations

In next-generation wireless networks, unmanned aerial vehicle (UAV) networks are emerging as a promising solution for ultra-reliable low-latency communication (URLLC). A key challenge in millimeter-wave UAV networks is ensuring that mobile users are always in line-of-sight (LoS) coverage, since the current snapshot-based trajectory planning approach does not consider the mobility of the users during the decision interval, resulting in disastrous LoS gaps. For continuous coverage verification, standard uniform sampling is too computationally expensive, as it would need a large number of samples to estimate rare failure events that have latencies that are not suitable for real-time requirements. In this work, …

Photoluminescent Studies Of Metal-Organic Frameworks For Small Molecule Sensing, Julia T. Dematteo

USF Tampa Graduate Theses and Dissertations

Metal–organic frameworks (MOFs) are highly tunable nanoporous materials with broad applications in gas separation, drug delivery, light harvesting, and chemical sensing. The incorporation of photoactive components into MOFs enables the development of optical sensors capable of detecting a wide range of analytes. Porphyrins and metalloporphyrins are particularly attractive for such applications due to their high molar absorptivity, strong emission quantum yields, and versatile interactions with chemical species.

In this work, two porphyrinic MOFs, PCN-222 and MOF1(Cd), are investigated. Both frameworks utilize tetrakis(4-carboxyphenyl)porphyrin (TCPP) as the organic linker, imparting optical functionality to the materials. Integration of TCPP into the framework induces …

Risk-Aware Deep Reinforcement Learning In Portfolio Optimization, Shimin Zhang

USF Tampa Graduate Theses and Dissertations

Portfolio optimization is important in investment, with the purpose of maximizing profit whilemanaging risk. The traditional portfolio optimization method, Modern Portfolio Theory, is based on the assumption that returns are normally distributed, which is always violated in practical financial market settings. In addition, traditional approaches to portfolio optimization are mostly designed as static optimizations and fail to account for the nonlinearity and non-stationarity of contemporary financial markets. Recent developments in deep reinforcement learning have attracted considerable attention for treating portfolio optimization as a sequential decision-making process. Traditional portfolio optimization methods usually evaluate risk based on the variance of the returns. …

Movement Ecology And Population Dynamics Of Box-Nesting Black-Bellied Whistling-Ducks (Dendrocygna Autumnalis) In Louisiana, Katie E. Miranda

LSU Master's Theses

Understanding demographic and behavioral processes underlying population dynamics is essential for effective wildlife management, especially for species subject to harvest pressure or experiencing population shifts. Black-bellied Whistling-Ducks (Dendrocygna autumnalis) are currently undergoing a notable northward range expansion and are increasing in abundance across the southeastern United States, raising management concerns for both ecological and economic reasons. As a result, there is growing public support for a special Black-bellied Whistling-Duck-specific harvest season, yet little research has been conducted on this species, particularly within newly-established populations. To address this gap, I conducted this study with the goal of evaluating seasonal …

Learning Trajectories Of Online Batch Selection Methods, Luke Green

Theses and Dissertations

Modern deep neural networks achieve strong performance on large-scale datasets, but often require substantial training time. Online batch selection methods seek to reduce this cost by updating models on informative subsets of each batch rather than on all available examples. Recently introduced methods leverage teacher models and report substantial speedups, particularly in noisy-label settings. However, comparisons are often based on the number of epochs required to reach a target test accuracy, a coarse metric that is sensitive to implementation details and may obscure important differences in learning dynamics. In this thesis, we implement several online batch selection methods in a …

Advancing Data Usability, Activity Modeling, And Stability Optimization In Computational Enzyme Design, Spencer Gardiner

Theses and Dissertations

A grand challenge of computational biology is to computationally design, in a single pass, a protein sequence that catalyzes an arbitrary chemical reaction at a high rate under specified conditions [1]. This work details advances in three essential areas on the path to that goal: data quality, activity prediction and modeling, and stability optimization. The structure and implementation of the Allotrope Simple Model (a FAIR data format for many scientific instruments) was examined in [2], setting the stage for training deep learning models on high-quality experimental datasets from diverse sources. In [3], the limits of physics-based and deep learning tools …

Beyond Single Images: A Comprehensive Benchmark For Album-Level Vision-Language Understanding, Shawn Huang

Theses and Dissertations

Automatic album organization has been studied extensively over the past decades due to significant progress in digital photography. Recent Vision-Language Models (VLMs) have shown strong performance on multi-image understanding, making them natural candidates for automating album organization workflows. While VLMs’ abilities in multi-image understanding have been widely studied, their performance on album organization remains underexplored. To bridge this gap, we introduce AlbumBench, the first comprehensive benchmark for automatic album organization. Specifically, we (1) define album organization tasks as photo selection for album-specific user objectives, photo rating according to how well user intents are fulfilled, and album-specific photo grouping given a …

Van Der Waals Interactions Through The Lens Of 2d Materials With Emphasis On Many-Body Effects, Thi Xuan Diem Dang

USF Tampa Graduate Theses and Dissertations

This dissertation addresses the long-standing challenge of accurately and efficiently modeling long-range electron correlations, with a particular focus on van der Waals (vdW) interactions. These interactions play an essential role in predicting the structure, stability, and functionality of molecular systems and condensed matter materials. Although Density Functional Theory (DFT) has become a robust cornerstone of computational materials science, traditional formulations of the theory frequently fail to capture dispersive interactions, leading to significant errors in determining interlayer distances, binding energies, and stacking configurations. These limitations become particularly pronounced in layered materials and vdW heterostructures (HSTs), where weak long-range correlations are sensitive …

Galaxy Morphology Classification Using Deep Learning, Dipanwita Kundu Roy

Master’s Dissertations

Galaxy morphology is the study of the shape and visual appearance of galaxies, such as spiral, smooth, edge-on, and other morphological types. Morphological classification plays an important role in understanding how galaxies form and evolve over cosmic time. Most existing machine learning approaches for galaxy morphology classification rely solely on RGB galaxy images, which primarily capture spatial information and lack the physical spectral context of galaxies. In contrast, astronomical spectral datacubes contain rich information across multiple wavelengths, providing insights into the internal and physical properties of galaxies. However, such spectral observations are available for only a limited number of objects. …

Relay Selection And User Scheduling In Reconfigurable Intelligent Surface Assisted Millimeter-Wave D2d Communication, Lakshmikanta Sau

Doctoral Theses

Reconfigurable intelligent surface (RIS) assisted millimeter wave (mmWave) device to device (D2D) communication has recently been proposed as a viable solution to support the overwhelming data traffic in fifth-generation (5G) and beyond wireless networks. However, due to the substantial propagation and penetration losses of mmWave, a direct line of sight (LoS) link between a pair of proximity devices is required for effective communication. Static obstacles like trees and buildings can easily obstruct the direct LoS connectivity between a device pair. In such cases, RISs help to establish an indirect LoS link between an obstructed device pair by reflecting the signals …

Vibrations Of Tapered Beam Via The Exterior Matrix Method, Simranjit Kaur

Student Theses and Dissertations

Cell phone towers, utility poles, and traffic signal poles all use hollow, tapered beams as their main structural element. Because these structures are tall and exposed to wind forces, understanding their vibration behavior is important for ensuring stability and safety. We will use the Exterior Matrix Method to analyse a single beam, which can be used to analyse compound structures involving tapered beams. First, we find the system of four equations satisfied by the tapered beam, which can be converted to a 4 x 4 matrix. Then we find the exterior matrix, which is a 6 x 6 matrix, corresponding …

The Ecotoxicological Implications Of Switching From Fluorescent To Light Emitting Diode Lighting For Zooplankton Culturing And Whole Effluent Toxicity Testing, Orithea Z. Regn

Student Theses and Dissertations

In the global transition toward light emitting diode (LED) lighting, an understudied area is the impact on the culturing and testing of Whole Effluent Toxicity (WET) testing organisms, more specifically, the zooplankton species. Without comparison to the current lighting used for culturing and testing, it is unknown if there will be an impact on performance, which could affect effluent, regulatory, and product safety decisions. This dissertation investigated if culturing and reference toxicity testing using sodium chloride for Ceriodaphnia dubia, Daphnia magna, and D. pulex under LED lighting was comparable to fluorescent. Comparisons were made between two laboratories and by season, …

American Sign Language Recognition And Analysis Using Deep Learning, Saurabh Kumar Soni

Master’s Dissertations

In this work I build a system that recognizes isolated American Sign Language (ASL) words, and I use it to ask one fairly direct question: when training data is scarce, is it better to look at the video pixels or at the geometry of the signer’s body? To find out, I train two very different models on exactly the same clips. The first is appearance-based. Every frame is run through standard preprocessing and a ResNet50 backbone pre-trained on ImageNet, which turns it into a 2048-dimensional feature vector, and a Bidirectional LSTM then reads that sequence over time. The second model …

Enhanced Embedding For Multimodal Medical Visual Question And Answering, Akash Suna

Master’s Dissertations

Visual Answering of questions in the field of Medical which is called as (VqA) has grown as a dominant area of research that fuse processing of natural language and vision of computer often known as CV or NLP to assist in medical decision-making. However, effective multimodal fusion between medical images and clinical questions remains a significant challenge. This thesis examines the application of the Perceiver IO architecture as an efficient multimodal aggregator for medical VQA. The work has been carried out in multiple directions. First, a classification-based framework is developed by combining Vision Transformer (ViT) and ClinicalBERT alongside a Perceiver …

An Empirical Study Of Rlvr Fine-Tuning For Mathematical Problem Solving In Llms, Rashmi Konnur

Master’s Dissertations

Large language models have shown immense improvement in coding and math performances thanks to reinforcement learning boosted algorithms. However, its true impact on broadening the reasoning and analytical capacities of an LLM is still contended. In this dissertation, we outline the foundations of Large Language Models, and delve into Reinforcement Learning with Verifiable Rewards (RLVR). We discuss various strategies to efficiently manipulate memory during a fine tuning update. We finally perform RLVR fine-tuning techniques on different models with varied use cases and compare their performances, which corroborate the efficiency of RLVR.

Photophysical Behavior Of Ruthenium-Based Complexes: A Study Of Ligand Variation And Encapsulation In Metal-Organic Frameworks, Delaney S. Sellers

USF Tampa Graduate Theses and Dissertations

Ruthenium (II) polypyridyl complexes have emerged as a prominent class of compounds in inorganic photochemistry due to their structural robustness, reversible redox chemistry, and tunable excited-state dynamics. Central to the investigation of these systems is the ability to manipulate the triplet metal-to-ligand charge transfer state (3MLCT). In the context of photoactivated chemotherapy (PACT), the strategic deactivation of this MLCT state to populate a dissociative triplet ligand field state (3LF) can enable the controlled, light-triggered release of caged cytotoxic agents. This work focuses on understanding the caging mechanism of various ligands to optimize the eventual delivery of cytotoxic ligands. In this …

Development Of A Coupled Hydrological-Nitrogen Modeling Framework For Urban Stormwater Pond Analysis, Tione Grant

USF Tampa Graduate Theses and Dissertations

Urbanization impacts the hydrological and biogeochemical processes that regulate water quality in downstream ecosystems, leading to increased stormwater volume and nitrogen loads in receiving aquatic environments. Although stormwater ponds are designed as simple hydraulic storage basins, they function more like managed ecosystems within urban drainage systems, offering both hydraulic attenuation and passive nutrient treatment. However, nitrogen removal efficiency varies significantly and depends on interactions between hydrological and biogeochemical factors. This thesis details the development, calibration, validation, and sensitivity analysis of a coupled hydrological-nitrogen model framework for Aaran’s Pond, an urban stormwater pond in Hillsborough County, Florida.

A process-based water budget …

Trophic Structure Of Reef-Fish Assemblages In The Eastern Gulf Of Mexico Varies By Depth And Reef Type, Jessica L. Van Vaerenbergh

USF Tampa Graduate Theses and Dissertations

Globally, marine ecosystems are increasingly threatened by anthropogenic stressors, leading to reef degradation, changes in reef-fish assemblages, and altered trophic structure. In this study, I investigated reef-fish underwater visual survey data collected over a decade (2013-2023) from paired natural and artificial reefs in the eastern Gulf of Mexico. I applied the novel Normalized Reef Status Index (NRSI) to evaluate how biomass was distributed across the three trophic categories (i.e., Lower Trophic Position (LTP; 2.0-2.9), Central Trophic Position (CTP; 3.0-3.9), and Upper Trophic Position (UTP; ≥4.0). This index is a metric to assess the ecological status of reef systems by investigating …

Development And Evaluation Of The East Florida Coastal Ocean Model (Efcom), Orion B. Scharton-Witmer

USF Tampa Graduate Theses and Dissertations

Florida’s Atlantic coast has abundant resources in beaches, state parks, aquatic preserves, and wildlife refuges as well as densely populated coastal cities in low lying areas that are all subjected to natural disaster threats from the oceans. There is an urgent need for a high-resolution coastal ocean circulation model for this region. Based on the application of the Finite Volume Community Ocean Model (FVCOM), an East Florida Coastal Ocean Model (EFCOM) is developed, downscaling from the deep ocean, across the continental shelf, and into the estuaries. It employs an unstructured grid with horizontal resolution varying from 3 kilometers at the …

Efficient Multimodal Foundation Model Tuning For Hallucination Mitigation, Fei Zhao

ETDs from 2020-2029

Over the past few years, multimodal foundation models have achieved remarkable progress in perception and understanding. However, two challenges limit their reliability: (1) dependence on offline training, which in most real-world settings requires large volumes of labeled data and, as a result, hinders the model’s ability to adapt to new data or domains; (2) weak cross-modal grounding, which often leads to hallucinated content generation, producing descriptions that are linguistically fluent but inconsistent with the input visual evidence. This dissertation frames hallucination mitigation as an outcome of transitioning from fixed learning (static, offline fine-tuning) to adaptive, feedback-driven lifelong learning. By incorporating …