Physical Sciences and Mathematics Commons^™

K-Nearest Neighbors Density-Based Clustering, Avory C. Bryant

Theses and Dissertations

Traditional density-based clustering approaches rely on a distance-based parameter to define data connectivity and density. However, an appropriate value of this parameter can be difficult to determine as it is highly dependent on the underlying distribution of the data. In particular, distribution parameters affect the scale of inter-group distances (e.g., variance); this dependence leads to a well-known inability to simultaneously detect clusters at varying levels of density. In this work, connectivity and density are defined according to the rank-order induced by the distance metric (i.e., invariant to the expected scale of the distances). Connectivity by k-nearest neighbors and density by …

Applied Machine Learning In Extrusion-Based Bioprinting, Shuyu Tian

Theses and Dissertations

Optimization of extrusion-based bioprinting (EBB) parameters have been systematically conducted through experimentation. However, the process is time and resource-intensive and not easily translatable across different laboratories. A machine learning (ML) approach to EBB parameter optimization can accelerate this process for laboratories across the field through training using data collected from published literature. In this work, regression-based and classification-based ML models were investigated for their abilities to predict printing outcomes of cell viability and filament diameter for cell-containing alginate and gelatin composite hydrogels. Regression-based models were investigated for their ability to predict suitable extrusion pressure given desired cell viability when keeping …

Fabrication Of Metal-Silicon Nanostructures By Reactive Laser Ablation In Liquid, Eric J. Broadhead

Theses and Dissertations

Metal-silicon nanostructures are a growing area of research due to their applications in multiple fields such as biosensing and catalysis. In addition, silicon can provide strong support effects to metal nanoparticles while being more cost effective than traditionally used supports, like titania. Traditional wet-chemical methods are capable of synthesizing metal-silicon nanostructures with a variety of composition and nanoparticle shapes, but they often require high temperatures, toxic solvents, strong reducing agents, or need capping agents added to stabilize the nanoparticles. Laser processing is an emerging technique capable of synthesizing metal-silicon composite surfaces that offers a faster, simpler, and greener synthesis route …

Detection Of 2,4,6-Trinitrotoluene Using A Colorimetric Gold Nanoparticle Air Cassette Filter, Andrea I. Ferrer Vega

Theses and Dissertations

Trinitrotoluene (TNT) is an explosive commonly used during military and terrorist activities. Current methods to identify this compound require sampling, transport and analysis at a forensic lab using analytical instrumentation. However, on-site detection is needed to assist efforts to prevent detonation. Gold nanoparticles have been used as sensors throughout the years due to their versatility and surface enhanced Raman scattering properties in the presence of an analyte and low limits of detection. By taking advantage of the Meisenheimer complex that TNT forms in the presence of amines, it is possible to determine its presence at picogram levels. Subsequently, adhering amine …

Evaluation Of Five Methods To Develop Latent Prints On Thermal Paper, Jenna Pregent

Theses and Dissertations

Forensic latent fingerprint laboratories determine the proper techniques for fingerprint visualization based on the substrates upon which they are deposited. Typical forensic analysis of thermal paper evidence involves the application of ninhydrin and/or 1,2-indanedione dissolved in a polar solvent. However, polar solvents create an undesirable reaction with the thermal paper’s internal properties and often lead to discoloration of the evidence. When this occurs, not only are the fingerprints less likely to be visible due to the loss of contrast, but the evidentiary print on the receipt may be lost entirely. This research sought to compare five development methods to determine …

Nebulizer-Based Systems To Improve Pharmaceutical Aerosol Delivery To The Lungs, Benjamin M. Spence

Theses and Dissertations

Combining vibrating mesh nebulizers with additional new technologies leads to substantial improvements in pharmaceutical aerosol delivery to the lungs across therapeutic administration methods. In this dissertation, streamlined components, aerosol administration synchronization, and/or Excipient Enhanced Growth (EEG) technologies were utilized to develop and test several novel devices and aerosol delivery systems. The first focus of this work was to improve the poor delivery efficiency, e.g., 3.6% of nominal dose (Dugernier et al. 2017), of aerosolized medication administration to adult human subjects concurrent with high flow nasal cannula (HFNC) therapy, a form of continuous-flow non-invasive ventilation (NIV). The developed Low-Volume Mixer-Heater (LVMH) …

Computational Study Of Radical Cation Rearrangements, Mi'kayla D. Word

Theses and Dissertations

A radical cation is a molecule that has one unpaired electron that holds a positive charge. The unpaired electron within a radical cation causes the molecule to be reactive. The high reactivity of these species allows for radical cations to be commonly studied experimentally using mass spectrometry and other multi-mass imaging techniques. However, these methods often cannot resolve the reaction mechanisms for these fast reactions. Specifically, radical cation rearrangement mechanisms are particularly unresolved within experiments. For this reason, radical cation rearrangements are computationally investigated to explain complex reaction pathways for processes to understand reactions leading to the initiation of detonation …

A Deep Learning U-Net For Detecting And Segmenting Liver Tumors, Vidhya Cardozo

Theses and Dissertations

Visualization of liver tumors on simulation CT scans is challenging even with contrast-enhancement, due to the sensitivity of the contrast enhancement to the timing of the CT acquisition. Image registration to magnetic resonance imaging (MRI) can be helpful for delineation, but differences in patient position, liver shape and volume, and the lack of anatomical landmarks between the two image sets makes the task difficult. This study develops a U-Net based neural network for automated liver and tumor segmentation for purposes of radiotherapy treatment planning. Non-contrast simulation based abdominal CT axial scans of 52 patients with primary liver tumors were utilized. …

Grief, Loss, And Climate Change: Validation Of A Solastalgia Scale, Claire Luce

Theses and Dissertations

Climate change has been identified as a defining issue of this century (United Nations, n.d.). Climate change impacts human wellbeing including mental health. While much research has focused on the way that the effects of climate change cause increases in common mental disorders, mental health is not just the absence of these disorders (World Health Organization, 2014). Non-pathologized mental health responses to climate change, such as the grief and loss that results from climate change impacts, are a growing consideration for researchers. Solastalgia, or the distress experienced in the absence of the solace once provided by the environment in the …

Continual Learning For Multi-Label Drifting Data Streams Using Homogeneous Ensemble Of Self-Adjusting Nearest Neighbors, Gavin Alberghini

Theses and Dissertations

Multi-label data streams are sequences of multi-label instances arriving over time to a multi-label classifier. The properties of the data stream may continuously change due to concept drift. Therefore, algorithms must adapt constantly to the new data distributions. In this paper we propose a novel ensemble method for multi-label drifting streams named Homogeneous Ensemble of Self-Adjusting Nearest Neighbors (HESAkNN). It leverages a self-adjusting kNN as a base classifier with the advantages of ensembles to adapt to concept drift in the multi-label environment. To promote diverse knowledge within the ensemble, each base classifier is given a unique subset of features and …

Equations Of State For Warm Dense Carbon From Quantum Espresso, Derek J. Schauss

Theses and Dissertations

Warm dense plasma is the matter that exists, roughly, in the range of 10,000 to 10,000,000 Kelvin and has solid-like densities, typically between 0.1 and 10 grams per centimeter. Warm dense fluids like hydrogen, helium, and carbon are believed to make up the interiors of many planets, white dwarfs, and other stars in our universe. The existence of warm dense matter (WDM) on Earth, however, is very rare, as it can only be created with high-energy sources like a nuclear explosion. In such an event, theoretical and computational models that accurately predict the response of certain materials are thus very …

Modeling Vegetation Effects On Barrier Island Evolution, Eric W. Schoen

Theses and Dissertations

Barrier islands play a significant role in protecting coastlines and harboring coastal habitats. In an effort to study and better understand the evolution of barrier island systems, a cellular model capturing various meteorological and environmental processes is proposed. Erosion due to wind, gravity, and marine processes are coupled with plant population effects. We demonstrate the inhibition of plant cover on sediment mobility, island migration, and erosion in the presence of sea level rise.

Parametric, Nonparametric, And Semiparametric Linear Regression In Classical And Bayesian Statistical Quality Control, Chelsea L. Jones

Theses and Dissertations

Statistical process control (SPC) is used in many fields to understand and monitor desired processes, such as manufacturing, public health, and network traffic. SPC is categorized into two phases; in Phase I historical data is used to inform parameter estimates for a statistical model and Phase II implements this statistical model to monitor a live ongoing process. Within both phases, profile monitoring is a method to understand the functional relationship between response and explanatory variables by estimating and tracking its parameters. In profile monitoring, control charts are often used as graphical tools to visually observe process behaviors. We construct a …

Ligand Effects On Electronic, Magnetic, And Catalytic Properties Of Clusters And Cluster Assemblies, Dinesh Bista 9288522

Theses and Dissertations

Ligands commonly protect metallic clusters against reacting with outside reactants. However, ligands can also be used to control the redox properties enabling the formation of super donors/acceptors that can donate/accept multiple electrons. This thesis focuses on how the ligands can be used to control the electronic and magnetic features of clusters and ligand stabilized cluster-based assemblies, leading to nano pn junctions with directed transport, the possibility of light-harvesting, and catalysts for cross-coupling reactions. The thesis addresses three distinct classes of clusters and their applications. The first class of cluster “metal chalcogen clusters” is the central idea of the thesis focused …

Reliable And Interpretable Machine Learning For Modeling Physical And Cyber Systems, Daniel L. Marino Lizarazo

Theses and Dissertations

Over the past decade, Machine Learning (ML) research has predominantly focused on building extremely complex models in order to improve predictive performance. The idea was that performance can be improved by adding complexity to the models. This approach proved to be successful in creating models that can approximate highly complex relationships while taking advantage of large datasets. However, this approach led to extremely complex black-box models that lack reliability and are difficult to interpret. By lack of reliability, we specifically refer to the lack of consistent (unpredictable) behavior in situations outside the training data. Lack of interpretability refers to the …

Single Molecule Investigations Of Holliday Junction Binding Protein Ruva, Dalton Reed Gibbs

Theses and Dissertations

DNA breaks are inevitable as they mainly occur due to cells’ own reactive oxygen species (ROS). While DNA breaks can be single-stranded or double-stranded, the double-stranded DNA (dsDNA) breaks are more dangerous. If such damage is not repaired, it can lead to genetic instability and serious health issues including cancers. One way dsDNA breaks can be repaired is via a process called homologous recombination (HR), which involves several DNA-binding proteins. Therefore, to have a better insight into the repair mechanism and origin of repair defects, we need a better understanding of how these proteins interact with DNA itself and DNA …

The Rational Synthesis Of Lanthanum Manganite Nanomaterials For Magnetic Refrigeration Applications, Caitlin Hunt

Theses and Dissertations

Refrigeration systems for the cooling of commercial and residential buildings are a major drain on energy resources and source of greenhouse gas emissions. Magnetic refrigeration, which employs the magnetocaloric effect, has the potential to mitigate the energy drain and greenhouse gas emissions for air conditioning. In order to commercialize magnetic refrigerators, a material with a high magnetic entropy change, ΔS, near room temperature which is simple and inexpensive to create is needed.¹ Current options for this solid refrigerant are either expensive, difficult to synthesize, or have a ΔS or T_c (the Curie Temperature, where ΔS is maximized) below …

Mathematical Modeling Of Lung Inflammation: Macrophage Polarization And Ventilator-Induced Lung Injury With Methods For Predicting Outcome, Sarah B. Minucci

Theses and Dissertations

Lung insults, such as respiratory infections and lung injuries, can damage the pulmonary epithelium, with the most severe cases needing mechanical ventilation for effective breathing and survival. Furthermore, despite the benefits of mechanical ventilators, prolonged or misuse of ventilators may lead to ventilation-associated/ventilation-induced lung injury (VILI). Damaged epithelial cells within the alveoli trigger a local immune response. A key immune cell is the macrophage, which can differentiate into a spectrum of phenotypes ranging from pro- to anti-inflammatory. To gain a greater understanding of the mechanisms of the immune response in the lungs and possible outcomes, we developed several mathematical models …

Live Cell Biomass Tracking For Basic, Translational, And Clinical Research, Graeme Murray

Theses and Dissertations

Single cell mass is tightly regulated throughout generations and the cell cycle, making it an important marker of cell health. Abnormal changes in cell size can be the first indication of dysfunction in response to environmental stimuli such as cytotoxic drugs. Described here is the further development of high-speed live cell interferometry (HSLCI) to concurrently measure the changes in single cell mass of thousands of cells over time. Critically, the high-throughput nature of HSLCI provides realistic pictures of tumor heterogeneity. This throughput enabled HSLCI to correctly predict in vivo carboplatin sensitivity of three triple negative breast cancer patient derived xenografts, …

Bisphosohoglycertae Mutase: A Potential Target For Sickle Cell Disease, Anfal S. Aljahdali

Theses and Dissertations

Bisphosphoglycerate mutase (BPGM) is a part of the erythrocyte glycolysis system. Specifically, it is a central enzyme in the Rapoport-Leubering pathway, a side glycolytic pathway involved in the regulation of the concentration of the natural allosteric effector of hemoglobin (Hb), 2,3-bisphosphoglycerate (2,3-BPG). BPGM catalyses the synthesis and hydrolysis of 2,3-BPG through its synthase and phosphatase activities. The synthase activity is the main role of BPGM, while the phosphatase activity is low and is activated by the physiological effector, 2-phosphoglycolate (2-PG) with the latter mechanism poorly understood.

BPGM activity and 2,3-BPG levels in red blood cells (RBCs) have a significant role …

Methods For Developing A Machine Learning Framework For Precise 3d Domain Boundary Prediction At Base-Level Resolution, Spiro C. Stilianoudakis

Theses and Dissertations

High-throughput chromosome conformation capture technology (Hi-C) has revealed extensive DNA looping and folding into discrete 3D domains. These include Topologically Associating Domains (TADs) and chromatin loops, the 3D domains critical for cellular processes like gene regulation and cell differentiation. The relatively low resolution of Hi-C data (regions of several kilobases in size) prevents precise mapping of domain boundaries by conventional TAD/loop-callers. However, high resolution genomic annotations associated with boundaries, such as CTCF and members of cohesin complex, suggest a computational approach for precise location of domain boundaries.

We developed preciseTAD, an optimized machine learning framework that leverages a random …

Statistical Approaches For Estimation And Comparison Of Brain Functional Connectivity, Jifang Zhao

Theses and Dissertations

Drug addiction can lead to many health-related problems and social concerns. Functional connectivity obtained from functional magnetic resonance imaging (fMRI) data promotes a variety of fundamental understandings in such association. Due to its complex correlation structure and large dimensionality, the modeling and analysis of the functional connectivity from neuroimage are challenging. By proposing a spatio-temporal model for multi-subject neuroimage data, we incorporate voxel-level spatio-temporal dependencies of whole-brain measurements to improve the accuracy of statistical inference. To tackle large-scale spatio-temporal neuroimage data, we develop a computationally efficient algorithm to estimate the parameters. Our method is used to identify functional connectivity and …

Characterizing Community-Level Size Spectra In Piedmont Streams Of Virginia (Usa), Giancarlo Racanelli

Theses and Dissertations

Many aquatic communities demonstrate an inverse scaling relationship between average body mass and density. Using quantitative samples of macroinvertebrates and fishes, we modeled this relationship in three Piedmont streams where little empirical research has been conducted. The size spectra (SS) method, in which individuals are identified by size, not taxonomic identity, was used with linear regression to model density as a function of mass. Fish and benthic invertebrate samples were collected on simultaneous days during September, then used to develop community-level SS models (combined fish and invertebrate data) for each stream. Invertebrate samples were also collected from each stream in …

Topics In Design And Analysis Of Experiments: Calibration, Sequential Experimentation, And Model Selection, Christine Miller

Theses and Dissertations

Experiments are widely used across multiple disciplines to uncover information about a system or processes. Experimental design is a statistical technique devoted to the methodology of selecting the appropriate samples to aid in the subsequent analysis. We research three open problems in experimental designs regarding calibration, sequential experimentation, and model selection. First, we focus on calibration; the impact of experimental design choice on the performance of statistical calibration is largely unknown. We investigate the performance of several experimental designs with regards to inverse prediction via a comprehensive simulation study. Specifically, we compare several design types including traditional response surface designs, …

Improving Space Efficiency Of Deep Neural Networks, Aliakbar Panahi

Theses and Dissertations

Language models employ a very large number of trainable parameters. Despite being highly overparameterized, these networks often achieve good out-of-sample test performance on the original task and easily fine-tune to related tasks. Recent observations involving, for example, intrinsic dimension of the objective landscape and the lottery ticket hypothesis, indicate that often training actively involves only a small fraction of the parameter space. Thus, a question remains how large a parameter space needs to be in the first place — the evidence from recent work on model compression, parameter sharing, factorized representations, and knowledge distillation increasingly shows that models can be …

Bayesian Techniques For Relating Genetic Polymorphisms To Diffusion Tensor Images Of Cocaine Users, Tmader Alballa

Theses and Dissertations

Past investigations utilizing Diffusion Tensor Imaging (DTI) have demonstrated that cocaine use disorder (CUD) yields white matter changes. We proposed three Bayesian techniques in order to explore the relationship between Fractional Anisotropy (FA), genetic data, and years of cocaine use (YCU). CUD participants exhibit abnormality in different areas of the brain versus non-drug using controls, which is measured by DTI. This dissertation is motivated by a neuroimaging genetic study in cocaine dependence, which found that there were relationships between several genes such as GAD and 5-HT₂R and CUD subjects.

In the first chapter, there is background on the …

Bayesian Experimental Design For Bayesian Hierarchical Models With Differential Equations For Ecological Applications, Rebecca Atanga

Theses and Dissertations

Ecologists are interested in the composition of species in various ecosystems. Studying population dynamics can assist environmental managers in making better decisions for the environment. Traditionally, the sampling of species has been recorded on a regular time frequency. However, sampling can be an expensive process due to financial and physical constraints. In some cases the environments are threatening, and ecologists prefer to limit their time collecting data in the field. Rather than convenience sampling, a statistical approach is introduced to improve data collection methods for ecologists by studying the dynamics associated with populations of interest. Population models including the logistic …

Information Architecture For A Chemical Modeling Knowledge Graph, Adam R. Luxon

Theses and Dissertations

Machine learning models for chemical property predictions are high dimension design challenges spanning multiple disciplines. Free and open-source software libraries have streamlined the model implementation process, but the design complexity remains. In order better navigate and understand the machine learning design space, model information needs to be organized and contextualized. In this work, instances of chemical property models and their associated parameters were stored in a Neo4j property graph database. Machine learning model instances were created with permutations of dataset, learning algorithm, molecular featurization, data scaling, data splitting, hyperparameters, and hyperparameter optimization techniques. The resulting graph contains over 83,000 nodes …

Learning From Multi-Class Imbalanced Big Data With Apache Spark, William C. Sleeman Iv

Theses and Dissertations

With data becoming a new form of currency, its analysis has become a top priority in both academia and industry, furthering advancements in high-performance computing and machine learning. However, these large, real-world datasets come with additional complications such as noise and class overlap. Problems are magnified when with multi-class data is presented, especially since many of the popular algorithms were originally designed for binary data. Another challenge arises when the number of examples are not evenly distributed across all classes in a dataset. This often causes classifiers to favor the majority class over the minority classes, leading to undesirable results …