Data Science Commons^™

Quantifying Grain Size In Scanning Electron Microscopy Images, Katherine Hoffsetz

Discovery Day - Daytona Beach

This project explores advanced image analysis techniques to assess the microstructure of highly strained austenitic stainless steel. Utilizing Python imaging libraries such as scikit-image and OpenCV, we aim to extract precise measurements for grain size from scanning electron microscopy (SEM) images. These metrics will be examined against the computed grain sizes of the sample from electron backscatter diffraction measurements. By automating the extraction of grain size measurements from SEM images, this study contributes to steamlining the quality assurance/ quality control of industrially processed materials.

Learning Casual Structures From Aviation Accident Narratives Using Natural Language Processing And Graph-Based Knowledge Representation, Stephanie Ramsey, Katherine Hoffsetz, Madeline Gorman, Logan Lambeth

Discovery Day - Daytona Beach

Understanding the complex causal relationships underlying aviation accidents is critical for improving safety and preventing future incidents. However, much of this information exists in unstructured narrative reports, making large-scale analysis difficult. This project aims to automatically extract and model causal chains from National Transportation Safety Board (NTSB) accident narratives using a combination of traditional natural language processing (NLP) techniques, transformer-based architectures, and graph-based knowledge representation. Traditional NLP methods, including named entity recognition, dependency parsing, and rule-based pattern matching, will be used to identify structured cause–effect relationships. These approaches will be compared with transformer-based models, including a lightweight encoder for classification …

Sequential Causal Architecture For Multimodal Aviation Accident Prediction, Kaitlyn Cavanaugh, Isaac Morrison

Discovery Day - Daytona Beach

Aviation accidents are rarely the result of a single failure but rather from a complex causal chain of latent failures. While traditional data mining models often predict incident occurrence, they frequently overlook the sequential mechanics defined by known accident causation theoretical frameworks like the Swiss Cheese Model and the FAA's HFACS. This project addresses the need for interpretable, reliable, multi-stage forecasting by proposing a Sequential Causal Architecture that transforms theoretical causation models into a structured Directed Acyclic Graph (DAG) for multimodal accident causation chain prediction. Data from the NTSB and DOT is used and connected together in a meaningful way …

Learning Motion Primitive Selection And Environment Abstraction, Edison Alberto Martinez Samaniego, Natalie Alexander, Kaelyn Weddle

Discovery Day - Daytona Beach

Learning Motion Primitive Selection and Environment Abstraction Advanced Air Mobility (AAM) is emerging as a transformative solution for short and medium range transportation; however, it introduces an operational model that differs significantly from conventional aviation. AAM vehicles are expected to operate closer to populated areas, with increased autonomy, in dense urban and suburban environments. These settings present constrained maneuvering conditions which highlights the importance of maintaining safe operation under degraded flight conditions. Abnormal conditions may endanger onboard passengers, people on the ground, and surrounding infrastructure, making rapid detection and mitigation essential to prevent loss of control. Recent research has explored …

Modeling Aircraft Collision Risk Using Machine Learning And Traffic Density Data Ac, Jadia Renee Ewing, Alexander Van Baelan, Conrad Prisby, Rafal P. Smietana

Discovery Day - Daytona Beach

Air traffic congestion is an increasingly important factor in aviation safety as global flight activity continues to grow. This project investigates whether higher traffic density is associated with an increased risk of aviation incidents and identifies key contributing factors. Using publicly available flight (ADS-B) and incident (NTSB) data, we apply several machine learning models to analyze traffic patterns and predict risk. We begin with logistic regression to evaluate the relationship between density and incident probability, followed by decision trees to extract interpretable rules describing high-risk conditions. K-nearest neighbors (KNN) is used to examine similarity in traffic patterns among incident flights, …

Low-Rank Spectral Analysis For The Reddening Of The Seven Sisters Star Cluster, Eric Rodarte, Angelina Scalice, Madison Warner, Kevin Numbe

Discovery Day - Daytona Beach

The Pleiades, also known as the Seven Sisters, is a stunning star cluster located approximately 440 light-years from Earth. This vibrant assemblage of hot blue stars in the Taurus constellation can be admired with the naked eye or through binoculars during early autumn. In this presentation, we utilize spectral theory to measure the reddening in the Pleiades star cluster. To evaluate the impact of interstellar dust on reddening, we employ principal component analysis (PCA) on a matrix representing color indices from various photometric bands linked to the cluster’s photometric data. This dataset was obtained from VIZIER. Our PCA analysis of …

Low-Complexity Polynomial Ring Learning For Quantum Space Assets, Lola Torres

Discovery Day - Daytona Beach

Secure communication for space-based systems requires cryptographic methods that remain both reliable and efficient under strict computational constraints. This work investigates a low-complexity polynomial ring learning algorithm designed for quantum space assets, including satellite–ground communication systems. The project focuses on post-quantum cryptographic principles, where encryption and decryption rely heavily on repeated polynomial operations; this can be computationally expensive with constrained platforms. This is addressed with reformulating polynomial multiplication as a structured linear transformation on coefficient vectors. By representing these operations as matrices with a cyclic structure, the structure allows the use of the discrete Fourier transform (DFT); this will simplify …

Small Uas Detection: Threat Intelligence & Risk Management Project, Tyler Johnson

Discovery Day - Daytona Beach

The TRANSPORTATION SECURITY ADMINISTRATION / FEDERAL AIR MARSHAL SUAS DETECTION: THREAT INTELLIGENCE & RISK MANAGEMENT PROJECT addresses the emerging safety and security challenges posed by the rapid growth of small Unmanned Aircraft Systems (sUAS) in complex airspace environments. This study analyzed 92 days of sensor-captured Remote Identification (RID) data collected near Fort Lauderdale-Hollywood International Airport (FLL) to assess operational behaviors, aviation risk, and ground risk associated with drone activity. The primary objective of this research is to identify patterns of unauthorized or hazardous sUAS operations to enhance situational awareness and inform actionable risk-mitigation strategies. The analysis identified 335 flights from …

Ai-Driven Scheduling Algorithms For Private Aviation, Tayan Benson, Jessica Buskey, Gabriel Camacho, Caitlyn A. Gabrinowitz

Discovery Day - Daytona Beach

Private aviation scheduling is complex and dynamic, requiring frequent aircraft repositioning based on demand and operational constraints, unlike fixed commercial airline schedules. As fleets grow beyond 300 aircraft, traditional deterministic methods become too slow, leading to the use of approaches such as genetic algorithms, but neural network-based methods have not seen in-depth exploration. This project models aircraft scheduling as a network, where airports and flights form a graph. It explores advanced AI methods, including graph neural networks and spatio-temporal graph neural networks (STGNNs), to capture both network structure and time constraints. The goal is to generate efficient daily schedules from …

Predicting Remaining Useful Life Using Multivariate Time-Series Data, Anayah Smith, Victoria Gaibor

Discovery Day - Daytona Beach

Accurate prediction of Remaining Useful Life (RUL) is critical for enabling predictive maintenance, improving system reliability, and reducing operational costs in degrading systems. This project addresses the problem of modeling and predicting RUL using multivariate time-series sensor data from the NASA CMAPSS turbofan engine dataset, with a focus on understanding how predictive performance changes across datasets of varying complexity. The objective is to develop a reproducible machine learning pipeline that captures degradation patterns and produces reliable time-to-failure predictions. The approach includes data preprocessing, exploratory data analysis, feature engineering, dimensionality reduction, and model evaluation. RUL values are computed and capped to …

Dcat - Distributed Computing And Analysis Tool, Asher Zwickel, Jacob Burdge

Discovery Day - Daytona Beach

This project uses distributed computing to process and analyze large datasets related to cyber breaches and attacks. Its main goal is to find patterns between initial cyber incidents and what happens next. It looks at whether responses tend to escalate, calm down, or stay about the same over time. Understanding this helps explain how digital conflicts develop and whether they follow predictable paths. The project was built as part of university research and runs on custom software across a cluster of 17 Chromebooks. While the system can study many topics, it is currently focused on cyber activity. The software uses …

Stormtrack: A Regime-Aware Classifier-Router Architecture For Multi-Horizon Kp Index Forecasting, John Rendleman

Discovery Day - Daytona Beach

STORMTRACK: A Regime-Aware Classifier-Router Architecture for Multi-Horizon Kp Index Forecasting Current algorithms in operational space weather face extreme difficultly predicting the Kp geomagnetic index beyond 24 hours, a lead time that is critical for protecting high-frequency communications and infrastructure. Most regression models are optimized for quiet conditions, which dominate the data, leading to systematic underpredictions of storm events that cripple space infrastructure. Probabilistic approaches and physics-based numerical models also falter due to the same class imbalance plaguing standard regressors at multi-day lead times. The ICARUS 6 architecture addresses this by splitting the forecasting component into quiet and storm regimes, which …

Meeting The Moment With Ai-Employer Informed Education, Brent Terwilliger, John Faraca

Publications

As artificial intelligence transforms aviation, aerospace, and autonomy-related sectors, higher education must adapt to meet evolving workforce demands. This session shares emerging findings from a nationwide study led by Embry-Riddle Aeronautical University, focused on employer perceptions of AI adoption, responsible use, and workforce preparedness in domains including uncrewed systems, space systems, robotics, and advanced air mobility. Based on a structured survey and follow-up interviews, the presentation explores how organizations are using AI tools, from generative platforms to enterprise systems, and defining effective and inappropriate use in operational contexts. Participants will gain insight into critical concerns (e.g., data privacy, compliance, security, …

Scenarioxp: A Complete Scenario-Based Testing Framework For The Exploration And Exploitation Of Autonomous Vehicle Validation Scenarios, Quentin Goss

Doctoral Dissertations and Master's Theses

Today is an age of exciting emerging technology where cutting-edge research in autonomous vehicles (AVs) reduces the active human participation in driving and extends awareness beyond human limitations of perception and reaction, improving driving safety and quality of the user experience as a result. The ever-increasing complexity of these autonomous systems poses many challenges towards the validation and verification (V\&V) of these complex systems under time and resource constraints, as the use of artificial intelligence and also the intricacy of the operating environment means that these systems are also black-box and non-deterministic. Scenario-based V\&V testing of such systems, which involves …

Low-Complexity Structured Neural Networks And Their Usage In Image And Signal Processing, Adam Kuzmicki

Doctoral Dissertations and Master's Theses

Conventional neural networks face significant challenges due to high computational costs, large parameter counts, and reliance on backpropagation, which restricts their application in resource-constrained and real-time settings. To address these challenges, this thesis proposes three structured neural network (NN) architectures grounded in the theories of sparse and self-contained factorizations of transforms, with applications to image compression, reconstruction, classification, encryption, and also adaptive wideband multi-beam beamforming. The first neural network architecture, named DCTrix-Net, replaces conventional spatial con- volution with highly sparse factorization of the discrete Cosine transform (DCT) complemented by Toeplitz-structured weight initialization, achieving at least 97% FLOP reduction over CNNs, …

Information Theory Analysis Of The Solar Wind Magnetic Structures For Space Weather Prediction, Katherine Holland

Doctoral Dissertations and Master's Theses

Forecasting space weather at Earth is highly complicated, because of the limited measurements of the dynamic processes in the Sun that span multiple temporal, spatial, and energy-scales. The solar wind is a highly structured, multi-scale, evolving plasma and consists of coronal mass ejections (CMEs), stream interaction regions (SIRs), expanding flux tubes (Borovsky, 2008), and interplanetary magnetic field (IMF) discontinuities and fluctuations. The aim of this research is to improve our understanding of the evolution and dissipation of different scale-size solar wind magnetic structures as they move from the Sun-Earth Lagrange point 1 (L1) to Earth's bow shock and, ultimately, to …

Terrorism By The Numbers: Event And Structural Determinants Of Attack Outcomes, Claire Lebakken

Student Research Symposium (SRS)

This project aims to identify which event-level and structural covariates are most predictive of terrorism outcomes. Using the Global Terrorism Database (1970–2020), we examine whether fatalities, injuries, attack type, target type, and actor type, combined with national-level conditions, can reliably predict outcomes such as lone actor versus group involvement, attack method, target selection, and property damage. Event-level data are merged with World Bank Development Indicators and Freedom House scores to incorporate economic and governance contexts. After cleaning the data, creating dummy variables, and log-transforming skewed measures (e.g., GDP per capita), we apply logistic and multinomial logistic regression models to test …

Parameter Informed Reinforcement Learning For Vehicle System Identification, Nathan Schaff

Doctoral Dissertations and Master's Theses

Accurate system identification is essential for modeling and controlling vehicle dynamics. This dissertation explores the application of Parameter Informed Reinforcement Learning (PIRL) as a novel approach to system identification (SYSID). PIRL integrates prior system knowledge, such as physical parameters, into reinforcement learning (RL) frameworks to improve estimation accuracy. The study begins with an overview of traditional SYSID methods and then introduces PIRL as a modification of standard RL. The research applies PIRL to short-period aircraft dynamics, demonstrating its effectiveness in both offline and online learning frameworks. The dissertation then further explores PIRL’s utility in an indirect model reference adaptive control …

Global-Local Method For Poroelasticity Problems With Localized Pressure Effects, Hemantha Kunwar

Math Department Colloquium Series

In many poroelasticity applications, pressure effects are confined to a small region, making it inefficient and possibly unnecessary to solve the full system across the entire domain. Instead, we propose to solve the poroelasticity problem locally, where pressure effects are significant, and use a simpler linear elasticity model elsewhere. This creates a coupled elasticity–poroelasticity problem with transmission conditions. To solve this coupled problem, we propose a new non-intrusive global–local algorithm that iteratively solves the elasticity problem in the entire (global) domain and the poroelasticity problem only in a local domain, ensuring proper transmission conditions across the interface. This approach, which …

A Hybrid Data Assimilation Approach For Parameter Estimation In Dynamical Systems, Xuejian Li

Math Department Colloquium Series

In this talk, we present a hybrid data assimilation (DA) method that integrates continuous data assimilation (CDA) with particle filtering to estimate parameters in dynamical systems. Parameter estimation in such systems is particularly challenging because it involves both determining the parameters and estimating the often high-dimensional physical state. To address this difficulty, we decouple the estimation of states and parameters by employing CDA for state estimation and particle filtering for parameter estimation, with information exchanged alternately between the two. This hybrid framework leverages the strengths of CDA in handling high-dimensional state estimation and the efficiency of particle filters in estimating …

Leveraging Machine Learning And Causal Inference For Loan Default Prediction, Luca Guida

Doctoral Dissertations and Master's Theses

This research explores a systematic application of machine learning techniques combined with causal inference to predict loan defaults in peer-to-peer lending. Accurately forecasting loan defaults is crucial for mitigating financial risk and optimizing lending strategies. This analysis is based on multiple datasets of loan applications spanning over a decade, containing detailed financial and credit information about borrowers. Beginning with extensive Exploratory Data Analysis (EDA) coupled with scaling strategies, the research identifies key trends in loan performance across a large number of factors, such as interest rates or borrower creditworthiness, and one objective is to determine from the many available predictors …

A Comparative Study Of Neural Networks And Xgboost Models For Flight Time Prediction, Ioannis Paraschos, Taryn E. Trimble, Eshna Bhargava, Jake Klingler, Benjamin R. Nicolai

Beyond: Undergraduate Research Journal

Flight time prediction plays a crucial role in modern air travel, benefiting airlines and passengers alike. Accurate predictions enable airlines to optimize schedules, allocate resources effectively, and ensure passenger safety and satisfaction. In recent years, machine learning models, such as neural networks and XGBoost, have gained popularity for predicting flight times. This study aims to compare the performance of neural network and XGBoost models in predicting flight times, considering factors such as weather conditions, air traffic control, and aircraft performance. The results indicate that both models are effective, with XGBoost achieving slightly higher accuracy. However, neural networks offer advantages in …

Statistical Study Of Solar Wind Conditions Prior To Substorm Onsets, Luke H. Francis

Doctoral Dissertations and Master's Theses

Due to complex, multi-region, coupled plasma systems, auroral substorm onsets have been historically difficult to predict. The northward turning of the interplanetary magnetic field was considered the primary candidate as an external triggering mechanism for substorm onsets. However, that was later shown to be coincidental in nature. This study is motivated by recent multi-spacecraft observations that show how several magnetosheath jets at the bow shock were heavily correlated to substorm onsets, indicated by a strongly radial IMF interval. In the past, studies have looked at small samples of substorms in order to make large-scale predictions. However in this study, a …

Applying Software Engineering Black-Box Methods For Testing Machine Learning Models, Timothy Elvira

Doctoral Dissertations and Master's Theses

This dissertation proposes researching an approach to incorporate and align Software black-box testing methods into Machine Learning (ML) applications, specifically in the context of computer vision models. Typically, testing methods within Software Engineering (SE) encompass a range of test types that assess levels of a software system, such as Unit, Integration, Functional, and System testing [1]. The testing spectrum offers two perspectives on the system: black-box, where the system’s code is hidden, and white-box, where the system's code is exposed for testing. Software Quality pairs testing with requirements, in a many-to-one relationship, to ensure proper validation of the software system. …

Quantifying The Transfer Effectiveness Of An Artificial Intelligence-Based Simulator Pre-Training Program For Student Pilots, Ryan Guthridge

Journal of Aviation/Aerospace Education & Research

Since the airline pilot shortage was initially studied in 2016, the pilot hiring model has been significantly impacted, with airlines hiring qualified pilots at unprecedented rates. The COVID-19 pandemic has slowed this hiring rate, however it is expected that airline hiring will soon increase to a rate higher than initially expected (Bureau of Transportation Statistics, 2022). With this dynamic, certified flight instructors are often the most qualified recruits for airlines, due to the number of hours and experience they have gained in the flight training organization. In turn, certified flight instructors are in short supply for flight training organizations worldwide. …

A Mathematical Model On The Temporal Dynamics Of Aviation Competitive Pricing, Tichaona Chikore,, Farai Nyabadza,

Journal of Aviation/Aerospace Education & Research

This study investigates the competitive dynamics of airport pricing using U.S. airport data to validate the findings. It employs linear and nonlinear ordinary differential equation models to analyze the influence of competitive interactions and internal factors on pricing decisions. The methodology involves parameter estimation via optimization techniques and quantile regression to capture heterogeneity across market segments. Mathematical analysis and simulation results show that if competitive coupling coefficients are low then there is weak competitive influence on pricing, with airports’ pricing largely driven by internal factors. Also, if the adjustment rates exhibit consistency across airports then internal dynamics are dominant in …

Exploring Machine Learning, Feature Engineering, And Explainability To Constrain Spica’S Apsidal Constant Through Mesa Simulations, Hannah C. Woodruff

Doctoral Dissertations and Master's Theses

Spica (α-Virginis) is a notable binary star system located in the constellation of Virgo and offers valuable insights into stellar interiors and dynamics. Within binary systems, gravitational forces between the two stars cause minor distortions that alter their orbital motion. The steady rate of this alteration is known as the apsidal constant, which provides key information about a star’s internal structure and its evolutionary state. Traditionally, stellar environments like Spica are studied using simulations, such as MESA (Modules for Experiments in Stellar Astrophysics). These simulations allow researchers to explore various aspects of stellar behavior through the entire evolution of the …

Localized Collocation Meshless Method For Modeling Transdermal Pharmacokinetics In Multiphase Skin Structures, Eduardo Divo

Math Department Colloquium Series

The human skin has a complicated structure with many multi-scale, biophysical effects impacting the propagation of skin-injected substances, such as partitioning, metabolic reactions, adsorption and elimination. An extended version of Fick’s second law governing the process of the compound diffusion in various skin layer is employed in the current work by considering the conservation of mass of the substance and the metabolic reaction of the substance in viable skin. Additionally, a model assuming linear coupling between the substance concentrations that are bound and unbound with blood was developed. Using such a model, a set of coupled partial differential equations are …

Resource Optimization For Air Mobility Under Emergency Situations, Yongxin (Jack) Liu

Math Department Colloquium Series

This project aims to improve air traffic management in emergencies. We first developed a GRU neural network to forecast weather-related airport capacity constraints using historical data, underscoring the value of real-time data analysis. We then optimized emergency evacuation air travel using Particle Swarm Optimization, demonstrating the ability to quickly aggregate evacuation flight resources cost-effectively. Finally, we provided a hybrid model combining a genetic algorithm with a neural network for evacuation planning, we show that neural network can be integrated accelerate genetic algorithms for efficient and performance assured system optimization.

Transfer Learning In The Era Of Foundational Models: Application To Diagnosis In Rheumatology, Prashant Shekhar

Math Department Colloquium Series

Problems with current synovitis grading procedures

• There has been a lack of reliability in grading these images in the medical community due to a lack of universally accepted diagnostic criteria [Momtazmanesh et al., 2022]
• The human/machine variability creates an additional challenge in an efficient automated scoring system [Ranganath et al., 2022]
• There is a lack of consistency between doctors in grading these images [Momtazmanesh et al., 2022]