Statistics and Probability Commons^™

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 …

Predicting Passenger Demand On National Flights Departing From Hartsfield-Jackson Atlanta International Airport (Atl) In 2024, Brooklyn Gossett, Ana Yu Wen

Discovery Day - Daytona Beach

The aviation industry relies heavily on accurate demand forecasting to guide critical decisions regarding route planning, capacity management, and pricing strategy. Misjudging passenger demand can result in significant revenue loss and operational inefficiency, making it essential for airlines and analysts to identify the key drivers of flight patronage. This study investigates the factors that most significantly predict the number of passengers on domestic flights departing from Hartsfield-Jackson Atlanta International Airport (ATL) during the 2024 calendar year. Using passenger and route data sourced from the Bureau of Transportation Statistics (BTS) and the U.S. Department of Transportation (DOT), a multiple regression analysis …

A Compact Representation Of Oscillatory Limits Via Asymptotic Value Distributions, Ibrahim Arnous, Eric M. Rodarte, Chirag Kumar

Discovery Day - Daytona Beach

Classical limits describe asymptotic behavior through convergence to a single value, but many important oscillatory functions do not converge in this sense. Standard examples such as sin(𝑥) as 𝑥→∞ and sin(1/x) as x→0 instead display stable distributions of values over time. This project examines how such behavior can be described using a measure-theoretic framework, particularly through occupation measures and, in sequence-based settings, Young measures. The objective is to present this perspective in a clear and accessible way while introducing the Ansatz representation, a compact notation for recording the support and density of an asymptotic value distribution when the limiting measure …

Bounded Thunderstorm Tracking Of Lightning Events With Muon Detection, Skylar Wardlaw, Ainsley Helgerson, Maria Kaminska, Logan Velvet, Georgii Dubrov, Jackson Stewart, Aaron Jung, Nathaniel O’Hara, Amelia Koth, Nash Mcleod, Emaleth Wyckoff

Discovery Day - Daytona Beach

The muon is a high-energy particle that can be produced by cosmic rays and trigger upper-atmospheric particle cascades and energetic processes. It has been hypothesized that such cascades are at the onset of lightning. As such, muons serve as a valuable probe for investigating the underlying mechanisms of its initiation, whose full governing dynamics remain vastly unknown despite extensive research. Traditional approaches to lightning research involve simulations and observations of the discharge itself, but the role of high-energy particle interactions has yet to be fully constrained. The CosmicWatch Muon Detector design enables the detection of atmospheric muons through scintillation events, …

Statistical Methodologies For Count Time Series Analysis And Topological Data Analysis Of Medical Images, Yuhyeong Jang

Statistical Science Theses and Dissertations

This dissertation addresses two distinct topics related to count time series analysis and topological medical image analysis, respectively. The first part of the dissertation comprises an application of a count time series model to analysis of US monthly sex trafficking data and development of a new model for multivariate count data that exhibits serial dependence and overdispersion. By imposing a family of multivariate mixed Poisson distributions on the count random vector, the proposed model can accommodate a broad range of overdispersion as well as positive contemporaneous correlations. For maximum likelihood estimation, a computationally feasible EM-type algorithm is derived based on …

Ground To Roof Snow Load Ratio (Gr) Data Release, Brennan Bean, Cooper Nelson, Jesse Wheeler, Scout Jarman, Salam Adil Al-Rubaye, Marc Maguire

Browse all Datasets

This data release provides historical ground-to-roof snow load ratio (GR) datasets used for snow load research and model development. The release includes original referenced datasets, cleaned country specific datasets, and a master dataset that combines Canadian and United States datasets into a standardized format for research and engineering applications.

Algebraic And Topological Methods In Computational Neuroscience, Trong-Thuc Trang

Electronic Theses and Dissertations

Neural data is incredibly rich in combinatorial, topological, and geometrical information, reflecting the intricate shape and connectivity of neural firing patterns. To decipher these structures, neuroscience increasingly relies on advanced mathematical tools to analyze neural activity. Here we study (1) neural codes within the poset P_Code of neural codes and (2) the connectivity of neural population activity within the insular cortex when responding to interoceptive information. In (1), we establish combinatorial constructions for all upward covering relations based on what we call “isolated subsets” with supporting theorems and give a slight modification of the existing downward covering relations. We …

A Semantic Data Management Framework For Uncertainty Quantification In Synchrotron Diffraction Pattern Analysis, Ayorinde E. Olatunde, Ozan Dernek, Erika I. Barcelos, Roger H. French, Anirban Mondal

Student Scholarship

Advancements in technology have enabled scientists and engineers in the synchrotron science domain to generate high-dimensional data. To conduct a machine learning study on these data, an analyst undergoes multiple analytical stages, including dimensionality reduction, predictive modelling, and uncertainty quantification (UQ).

To ensure reproducibility in science, it is necessary to document the semantic relationships among the stages of the analysis to enable provenance, interoperability, and knowledge reuse. In this work, we present a Semantic Data Management (SDM) framework rooted in FAIR principles that provides ontologies for UQ workflows. We illustrate the application of UQ ontology in synchrotron diffraction pattern analyses …

Wavelet-Based Multiscale Analysis Of Cave Co₂ Concentration In Response To Short-Term High-Intensity Tourism Activities: Spatiotemporal Heterogeneity And Lag Characteristics, Mingda Cao, Wenwen Song, Yan Zhang, Jie Zhang, Zhiqiang Yao

International Journal of Speleology

High-intensity tourism activities can cause a significant increase in cave air CO₂ concentration, thereby affecting the cave micro-environment and secondary carbonate deposition. During the 2023 National Day Golden Week, high-frequency continuous monitoring of cave air CO₂ partial pressure (PCO_2(A)) and visitor numbers was conducted in Dawang Cave, Anhui Province. Wavelet transform and cross-correlation analyses were used to reveal the multi-scale response characteristics of CO₂ concentration to tourism activities. The results show that: (1) PCO_2(A) exhibited clear diurnal variations (higher during the day, lower at night) and decreased spatially with enhanced ventilation, controlled jointly by …

Criticality In A Heterogeneous Neutron Transport Rod Model, Samuel Kaleb Crowford

All Graduate Reports and Creative Projects, Fall 2023 to Present

This work studies the stochastic behavior of neutron populations in a one-dimensional rod model using Monte Carlo simulation. The first part of this project reproduces the computational results of Dumonteil, Horton, Kyprianou, and Zoia (2025) by independently implementing the Monte Carlo algorithm described in their article, with the asymptotic behavior of the first moment analyzed in relation to the dominant eigenvalue and adjoint eigenfunction of the neutron transport operator. The model is then extended to a heterogeneous setting by introducing a central region where fission is suppressed. A global expectation over initial positions and directions is used to estimate the …

Mapler: An R Package For Estimating The Impact Of Climate Change On Maple Syrup Production, Matthew T. White

All Graduate Theses and Dissertations, Fall 2023 to Present

Successful maple sap tapping depends on the freeze/thaw cycle (i.e., temperatures fluctuating above/below freezing) during the winter and spring. Climate change threatens to alter the timing and duration of the tapping season. This necessitates research into how maple sap tapping will be impacted by climate change in order to help maple syrup producers prepare for the future. We define a sap day as a day where the freeze/thaw cycle occurred. Using information climate scientists use to predict future temperatures, we calculate how many sap days could occur each year. We develop software to analyze these sap day calculations to determine …

When The Best-Fit Model Is Not Best: The Glass Slipper Fallacy And Latent Growth Mixture Modelling, Jonathan L. Chia, Markus Wettstein, Andree Hartanto

Research Collection School of Social Sciences

Despite the use of latent growth mixture modelling (LGMM) to study longitudinal changes, existing practices may inadvertently impede this very investigation. Although subgroup trajectories may theoretically differ in their structure (e.g., some subgroups being linear, some curvilinear), the current convention advocates overreliance on the baseline model to derive subsequent profile trajectories, which may obscure these structural differences. In this article, we provide a brief description of extant LGMM practices, after which we explicate the pitfalls of the current approach. Finally, we provide a principled approach for LGMM research moving forward. Specifically, we recommend specifying a set of theoretically plausible models …

Modeling Mean And Variability Of Anxiety In Ecological Momentary Assessment Data Using Mixed-Effects Location–Scale Models, Trenzy Odero

Electronic Theses and Dissertations

Ecological Momentary Assessment is a method of collecting repeated measures of people in real time within natural environments. This results in hierarchical data that has a significant amount of variation at the person level. The traditional linear mixedeffects models assume that the residual variance is constant, which might not be true when the residual variance varies among individuals as well as in time. This thesis uses mixed-effects location-scale (MELS) models to model the mean and variance of an EMA outcome together. By introducing the possibility of variability in residual variance within and across individuals and with covariates, the MELS framework …

Machine Learning For Predictive Energy And Emissions Modeling Of Vehicles And Power Grids In The United States, S M Tanvir Faysal Alam Chowdhoury

Dissertations

The environmental benefits of electric vehicle (EV) adoption depend on more than replacing internal combustion engine vehicles with electric powertrains. EV adoption reshapes electricity demand, interacts with regional generation mixes, and influences travel behavior and congestion, creating a coupled transportation-energy system in which vehicle and power-plant emissions must be evaluated together. This dissertation develops machine-learning frameworks for predicting energy consumption and emissions from vehicles and power grids under rising EV adoption. The first component forecasts grid emissions from EV charging. Using simulation data from NREL's Cambium database, a Prophet-based time-series framework predicts carbon dioxide, nitrous oxide, and methane emission rates …

Laplace Factor Models In High-Dimensional Data, Siqi Liu, Xuerong Meggie Wen, Akim Adekpedjou, Guangbao Guo

Mathematics and Statistics Faculty Research & Creative Works

Laplace factor models (LFMs) provide a heavy-tailed alternative to Gaussian factor models by representing high-dimensional observations through a low-rank common component and Laplace-distributed idiosyncratic errors. This paper develops an assumption-consistent finite-sample analysis of matrix concentration, covariance estimation, and Monte Carlo integration under this model. We first formulate the model with explicit dimensional, independence, covariance, and identifiability conditions. Standard matrix Laplace-transform and matrix Bernstein inequalities are then recalled with their precise applicability conditions. Because untruncated Laplace variables are neither almost surely bounded nor strongly log-concave, these standard results cannot be applied directly in the forms commonly used for bounded or Gaussian-like …

Learning Latent Structure In High-Dimensional Data Via Geometry And Graphs, Haozhe Chen

All Graduate Theses and Dissertations, Fall 2023 to Present

Modern datasets often contain many measured variables for each observation, such as gene-expression levels, brain activity signals, or features in tabular data. These data are also often noisy, meaning that useful patterns are mixed with measurement error or irrelevant variation. Although such datasets can appear complex, they are frequently represented by simpler hidden structures, such as trajectories, clusters, or relationships between observations. This dissertation develops methods for uncovering these hidden structures by learning geometric and graph-based representations directly from data. The first part introduces Functional Information Geometry, which represents local patterns in high-dimensional data using functional features and constructs a …

Unifying And Expanding Global And Local Variable Importance Methods For Explainable Machine Learning, Kelvyn K. Bladen

All Graduate Theses and Dissertations, Fall 2023 to Present

Machine learning methods are powerful analytical tools used across all scientific disciplines and many other fields of investigation for prediction and inference from diverse data sources. Despite their broad applicability, machine learning methods are often highly complex and difficult to interpret. Developing a greater understanding of which variables most influence a response is essential for increasing the interpretability of these models and supporting informed decision-making. This research focuses on improving how we evaluate the importance of these variables.

One common approach is to shuffle the values of a variable and see how much the model accuracy gets worse. Another approach …

Spatial Prediction Under Uncertainty: Methodological And Computational Advances In Bayesian Maximum Entropy, Kinspride K. Duah

All Graduate Theses and Dissertations, Fall 2023 to Present

Environmental decisions such as infrastructure design, water management, and snow load estimation depend on spatial data that are often incomplete or uncertain. In many cases, measurements are not exact values but ranges, reflecting limitations in data collection methods. Traditional mapping techniques typically simplify these uncertain measurements, which can lead to less accurate predictions. This dissertation introduces improved statistical tools for making spatial predictions when data are uncertain or partially known. By utilizing a framework called Bayesian Maximum Entropy (BME), this research demonstrates how exact measurements and range-based data can be combined in a mathematically consistent way. The work demonstrates that …

Community Detection In Bipartite Networks Using Bipartite Stochastic Block Models With Node-Level Covariates, Geraldine Elaine Percival

Dissertations

In this age, monumental webs of data demands for perpetual cultivation of ways to untangle these webs of information. One of the many curiosities is how to systematically group entities. When clustering, one avenue to take is ascertaining the interconnectedness between the data points, and gauging their influence to each other. This perspective is programmed to model the relationships between the data presented as a network. Many of the methods being used today are algorithm-based, which may pose limitations in understanding and explaining the uncertainty revolving around the data. Hence, it is proposed to steer towards a model-based approach that …

Evaluating Machine Learning Models On Classification Of Novel Cyber Attacks In The Healthcare Domain, Promise Ehimen

Dissertations, Theses, and Projects

The increasing adoption of the Internet of Medical Things (IoMT) has improved healthcare delivery through connected medical devices while simultaneously expanding the cybersecurity risks facing healthcare organizations. Although machine learning based intrusion detection systems have demonstrated high detection accuracy, their ability to respond reliably to previously unseen cyberattacks remains uncertain. This study investigated how a Neural Network model and a Logistic Regression model classified novel cyberattacks within the IoMT environment. The Neural Network and Logistic Regression models were both trained and tested using a subset of the CICIoMT2024 benchmark dataset. The Neural Network achieved 99.82% test accuracy and a 0.94 …