Digital Commons Network^™

Optimizing Neural Network Architecture Using Kernel Principal Component Analysis, Saige Simcox

Theses and Dissertations

Neural networks have proven to be powerful tools for modeling a wide range of problems across applications. However, one of the challenges in implementing a neural network model lies in determining the neural network architecture, i.e. the appropriate number of hidden layers and the number of neurons per hidden layer. It has been suggested that one way to determine the number of hidden layers is by using information on the variability captured by each principal component. In this research, we expand on this idea and propose a new approach to determine the neural network architecture for a multilayer perceptron used …

Synergetic And Orthogonal Switching In Dual Photochromic Metal-Organic Frameworks, Gina Rene Wilson

Theses and Dissertations

Cooperative behavior and orthogonal responses of two classes of coordinatively integrated photochromic molecules towards distinct external stimuli were demonstrated on the first example of a photo-thermo-responsive hierarchical platform. Synergetic and orthogonal responses to temperature and excitation wavelength are achieved by confining the stimuli-responsive moieties within a metal-organic framework (MOF), leading to the preparation of a novel photo-thermo-responsive spiropyran-diarylethene based material. Synergistic behavior of two photoswitches enables the study of stimuli-responsive resonance energy transfer as well as control of the photoinduced charge transfer processes, milestones required to advance optoelectronics development. Spectroscopic studies in combination with theoretical modeling revealed a nonlinear effect …

Mild Hydrothermal Crystal Growth Strategies For Complex Inorganic Metal Fluorides, Navindra Keerthisinghe

Theses and Dissertations

Inorganic metal fluorides have been investigated for their interesting magnetic and optical properties, their electrochemical behavior, and their ionic conductivity. Fluoride materials can accommodate metals in very low oxidation states, which can be advantageous for designing new magnetic materials. Specifically, mixed metal fluorides incorporating transition metal cations often exhibit interesting magnetic behavior, such as frustrated magnetism. However, many challenges are encountered during the synthesis of such fluoride materials, including the inability to perform reactions under ambient conditions and the involvement of toxic gases for fluorination during solid-state reactions. Therefore, solution growth techniques such as mild hydrothermal synthesis are currently utilized …

Predictive Filtering-Based Image Inpainting, Xiaoguang Li

Theses and Dissertations

Image inpainting is an important challenge in the computer vision field. The primary goal of image inpainting is to fill in the missing parts of an image. This technique has many real-life uses including fixing old photographs and restoring ancient artworks, e.g., the degraded Dunhuang frescoes. Moreover, image inpainting is also helpful in image editing. It has the capability to eliminate unwanted objects from images while maintaining a natural and realistic appearance, e.g., removing watermarks and subtitles. Disregarding the fact that image inpainting expects the restored result to be identical to the original clean one, existing deep generative inpainting methods …

Macro–Micro-Coupled Simulations Of Bead–Spring Breaking-Reforming Networks, Andrei Medved

Theses and Dissertations

In this work we investigate the dynamic behavior of bead-spring polymer solutions in viscoelastic fluids, which are essential in industries like materials science, biotechnology, and pharmaceuticals. The study leverages GPU-accelerated simulations and detailed modeling of polymer chain dynamics at the mesoscale, which enables efficient analysis of intricate fluid behaviors and the microscale dynamics of polymer chains. Additionally, the research examines the breaking-reforming dynamics of polymer chains, crucial for understanding phenomena such as shear thinning and thickening. The findings have broad applications, from improving inkjet printing and 3D printing technologies to developing new drug delivery systems and biocompatible materials. This work …

Generalizations Of The Graham-Pollak Tree Theorem, Gabrielle Anne Tauscheck

Theses and Dissertations

Graham and Pollak showed in 1971 that the determinant of a tree’s distance matrix depends only on its number of vertices, and, in particular, it is always nonzero. This dissertation will generalize their result via two different directions: Steiner distance k-matrices and distance critical graphs. The Steiner distance of a collection of k vertices in a graph is the fewest number of edges in any connected subgraph containing those vertices; for k = 2, this reduces to the ordinary definition of graphical distance. Here, we show that the hyperdeterminant of the Steiner distance k-matrix is always zero if …

Metal-Organic Frameworks For Global Environmental Challenges, Kyoungchul Park

Theses and Dissertations

Over the past few decades, there has been a demand for improving nuclear waste administration due to a significant increase in the utilization of nuclear energy. The driving force for proper nuclear waste disposal is evident when considering past nuclear catastrophes—such as Chernobyl and Fukushima—where radionuclide contamination of large bodies of water sparked public concern. In order to manage such radioactive species, the emerging areas of radionuclide sequestration, storage, separation, and sensing must be addressed in the upcoming years. This dissertation presents a potential candidate for addressing these challenges, called metal-organic frameworks (MOFs). Due to their porosity, high surface area, …

Controlling Properties Of Hybrid Materials: Leaching Kinetics And Electronic Structure, Anna Berseneva

Theses and Dissertations

Hybrid materials are a unique class of materials that merge the properties of organic and inorganic building blocks. These materials are at the frontier of material science due to their wide modularity and applicability in numerous fields, ranging from catalysis to nuclear waste treatment. Unprecedented tunability of physical properties via changes in organic, inorganic moieties and their connectivity, enables for hybrid materials to address current challenges in the materials’ applications improving energy landscape.

The presented work consists of six chapters where the synthesis and design of new hybrid materials, fundamental understanding of electronic behavior and its modulation in organic-inorganic materials, …

Representation Dimensions Of Algebraic Tori And Symmetric Ranks Of G-Lattices, Jason Bailey Heath

Theses and Dissertations

Algebraic tori over a field k are special examples of affine group schemes over k, such as the multiplicative group of the field or the unit circle. Any algebraic torus can be embedded into the group of invertible n x n matrices with entries in k for some n, and the smallest such n is called the representation dimension of that torus. Representation dimensions of algebraic tori can be studied via symmetric ranks of G-lattices. A G-lattice L is a group isomorphic to the additive group Zⁿ for some n, along with an action …

Modeling, Analysis, Approximation, And Application Of Viscoelastic Structures And Anomalous Transport, Yiqun Li

Theses and Dissertations

(Variable-order) fractional partial differential equations are emerging as a competitive means to integer-order PDEs in characterizing the memory and hereditary properties of physical processes, e.g., anomalously diffusive transport, viscoelastic mechanics and financial mathematics, and thus have attracted widespread attention. In particular, optimal control problems governed by fractional partial differential equations are attracting increasing attentions since they are shown to provide competitive descriptions of challenging physical phenomena. Nevertheless, variable-order fractional models exhibit salient features compared with their constant-order analogues and introduce mathematical difficulties that are not typical encountered in the context of integer-order and constant-order fractional partial differential equations.

This dissertation …

Advances In Chalcogenide Crystal Growth, Anna Berseneva

Theses and Dissertations

Chalcogenides are the cornerstone of the semiconductor and thermoelectric industries and are up-and-coming materials for superconductors, catalysis, and battery applications. Challenges in the synthesis of those materials emerge from the chalcogen’s volatility and the tendencies of chalcogenide to react with even trace quantities of oxygen. Nonetheless, many techniques have been applied to the growth of chalcogenide single crystals, which are convenient for structure determinations and intrinsic property measurements. This dissertation will detail synthetic strategies applied to the crystal growth of novel chalcogenide materials via molten flux preparation (Chapters 2, 4, and 5), hydrothermal synthesis (Chapter 1 and 3), and single-crystal-to-single …

Deconvolving Provenance Signatures Of Pliocene-Pleistocene Coastal Loess Deposits Along The Atlantic Margin Of Argentina, Will Robert Quanrud

Theses and Dissertations

Dust derived from southern South America provides critical nutrients for autotropic marine biota in the South Atlantic Ocean, such that fluxes of wind-blown sediment may ultimately affect the concentration of atmospheric CO2. Expansive sand, silt, and loess deposits are exposed in Central Argentina, recording millions of years of eolian processes; however, the provenance and transport pathways of this dust remains unclear. Coastal loess deposits ranging in age from Pliocene to upper Pleistocene are present along the Atlantic margin of Argentina and provide information on sediment transport pathways for the last 4 million years. Nine samples of loess were collected and …

Bayesian Methods In Analyzing The Diagnostic Accuracy\\ For Ordinal Ratings, Yun Yang

Theses and Dissertations

This dissertation focuses on ordinal classification ratings, which are commonly used in medical practice to assess the severity of a disease or condition. For example, a group of radiologists rate a set of mammograms and assign BI-RADS (Breast Imaging Reporting Data System) score for each mammogram. A Bayesian probit hierarchical model is first proposed to analyze this type of data. It links the ordinal ratings with both rater diagnostic skills and patient latent disease severity. Each rater diagnostic skills are quantified with two parameters, diagnostic bias and diagnostic magnifier. Patient latent disease severity is assumed to follow a different normal …

Object Classification, Detection And Tracking In Challenging Underwater Environment, Md Modasshir

Theses and Dissertations

The main contributions of this thesis is the applicability and architectural designs of deep learning algorithms in underwater imagery. In recent times, deep learning techniques for object classification and detection have achieved exceptional levels of accuracy that surpass human capabilities. However, the effectiveness of these techniques in underwater environments has not been thoroughly researched. This thesis delves into various research areas related to underwater environments, such as object classification, detection, semantic segmentation, pose regression, and semi-supervised retraining of detection models.

The first part of the thesis studies image classification and detection. Image classification is a fundamental process that involves assigning …

Electronic Structure Investigation Of Photoswitch-Embedded Organic Frameworks And External Electric Field Effects On Carbon Dioxide, Austin Hill

Theses and Dissertations

Metal- and Covalent-Organic Frameworks (MOFs and COFs, respectively) are chemically versatile systems that have diverse applications, ranging from drug delivery to energy storage. This versatility is due to the highly tunable nature of organic frameworks, which utilize metal clusters, chemically diverse metals, and organic ligands. Researchers have recently demonstrated how MOFs and COFs can be remotely tuned through the use of photochromic organic linkers and light (Martin, et al., 2022). The implications of these findings are significant and have numerous applications, primarily in optical switching and energy storage. However, the photochromic mechanism within the organic framework must be investigated, which …

Molecular And Hybrid Catalysts For Selective Hydrodeoxygenation Of Lignin Model Compounds, Jacob G. Tillou

Theses and Dissertations

Lignin is a highly aromatic, branched polymer and could be utilized as a renewable source of liquid fuel. The advent of “lignin-first” depolymerization processes have allowed for the high-yield isolation of oxygenated aromatic monomers from lignin. These oxygenated aromatic compounds are lower in energy and stability than their deoxygenated counterparts. Selective deoxygenation of these compounds would result in stable, higher value liquid fuels. Using molecular catalysts, selective deoxygenation, or hydrodeoxygenation (HDO), of lignin-derived compounds in the presence of hydrogen is possible.

Immobilization of these molecular catalysts on metal oxide surfaces eliminates the possibility of bimolecular decomposition and allows for easier …

Exiting Inflation With A Smooth Scale Factor, Harry Oslislo

Theses and Dissertations

The expectation that the physical expansion of space occurs smoothly may be expressed mathematically as a requirement for continuity in the time derivative of the metric scale factor of the Friedmann–Robertson–Walker cosmology. We explore the consequences of imposing such a smoothness requirement, examining the forms of possible interpolating functions between the end of inflation and subsequent radiation- or matter-dominated eras, using a straightforward geometric model of the interpolating behavior. We quantify the magnitude of the cusp found in a direct transition from the end of slow roll inflation to the subsequent era, analyze the validity several smooth interpolator candidates, and …

Unraveling The Structure And Star Formation History Of Polar Ring Galaxies, Kyle Lackey

Theses and Dissertations

Polar ring galaxies (PRGs) are a fascinating subgroup of early-type galaxies (ETGs) consisting of a central host galaxy - typically E/S0 - surrounded by a ring of gas, dust and stars orbitting perpedicular to the major axis of that host. In order to estimate physical properties and origins of five such objects taken from Whitmore's 1990 PRC catalog, we conduct a photometric study in optical and infrared wavebands. Data are taken from SDSS, GMOS-S, HST-WFPC2, and SST-IRAC. Each object was decomposed into a host galaxy bulge, host galaxy disk, and polar ring, and each component was fit with a Sersic …

Impaired Waters: New Treatments And Improved Methods For Pfas And Algae Toxins, Alexandria Forster

Theses and Dissertations

Per- and polyfluoroalkyl substances (PFAS) are high-profile environmental contaminants, many having long persistence in the environment and widespread presence in humans and wildlife. Following phase-out of perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA) in North America and restrictions in Europe, PFAS replacements are now widely found in the environment. While liquid chromatography (LC)-mass spectrometry (MS) is typically used for measurement, much of the PFAS is missed. To more comprehensively capture organic fluorine, we developed sensitive and robust methods using activated carbon adsorption, solid phase extraction, and combustion ion chromatography (CIC) to measure total organic fluorine (TOF) in industrial wastewaters, river …

Erlang-Distributed Seir Epidemic Models With Cross-Diffusion, Victoria Chebotaeva

Theses and Dissertations

We examine the effects of cross-diffusion dynamics in epidemiological models. Using reaction-diffusion dynamics to model the spread of infectious diseases, we focus on situations in which the movement of individuals is affected by the concentration of individuals of other categories. In particular, we present a model where susceptible individuals move away from large concentrations of infected and infectious individuals.

Our results show that accounting for this cross-diffusion dynamics leads to a noticeable effect on epidemic dynamics. It is noteworthy that this leads to a delay in the onset of epidemics and an increase in the total number of people infected. …

Statistical Inference Based On Elliptically Symmetric Distributions For Directional Data, Zehao Yu

Theses and Dissertations

Directional data arise in many scientific fields such as meteorology, oceanography, geology, zoology, and biomechanics. Geometrically, directional data lie in a spherical space. Although not in a spherical space, compositional data, such as microbiome data, can be mapped from a simplex to a spherical space via the component-wise square-root transformation. Other examples where compositional data emerge as a subject of interest include compositions of minerals in rocks, compositions of chemical mixtures, investment portfolios, and demographic composition of a population. This dissertation aims to develop inference procedures for analyzing directional data in general initially, with later focus shifted to the transformed …

Investigating Arctic Ocean Freshwater Content And Circulation In A Changing Climate, Emma Louise Hoffman

Theses and Dissertations

The Arctic Ocean is unique among the world’s oceans, characterized by its cold temperatures, salinity-driven vertical structure, and shallow shelf seas. It has warmed intensely since the 1980s at rates exceeding the global average. This Arctic Amplification drives variability in sea ice and salinity, affecting local circulation and exchange with other basins, leading to global changes in oceanic and atmospheric circulation and the biosphere. This thesis explores changing dynamics in the Arctic Ocean, in particular connecting modeled changes in liquid freshwater content (FWC) and its pathways with major ocean circulation patterns, atmospheric drivers, and large-scale climatic changes.

This research combines …

Global Well-Posedness Of Nonlocal Differential Equations Arising From Traffic Flow, Thomas Joseph Hamori

Theses and Dissertations

Macroscopic traffic flow models describe the evolution of a function ρ(t, x), which represents the traffic density at time t and location x according to a differential equation (typically a conservation law). Numerous models have been introduced over the years which capture the phenomenon of shock formation in which the solution develops a discontinuity. This presents difficulties from the standpoint of mathematical analysis, necessitating the consideration of weak solutions. At the same time, this undesirable mathematical behavior corresponds to unsafe driving conditions on real roadways, in which the heaviness of traffic may vary abruptly and dramatically. This thesis introduces and …

Assessing Multi-Annual Phytoplankton Community Compositions And Biomass Across Tidal Creek And Open Water Estuarine Habitats In South Carolina, Shannon Elizabeth Powers

Theses and Dissertations

The urbanization of coastal environments in recent decades has caused an accelerating increase in nutrient-rich runoff from the landscape. The presence of these excess nutrients in the aquatic environment can result in degradation of water quality and harmful algal blooms. In South Carolina (SC), coastal development continues to threaten the resiliency of salt marsh estuaries, especially in combination with other stressors like climate change and sea level rise. Estuaries are important nurseries and habitats for fisheries and ecotourism. The South Carolina Estuarine and Coastal Assessment Program (SCECAP) is an ongoing monitoring program that assesses the habitat condition along the coast …

Molecular Rotors For Quantifying Kinetic Effects Of Stabilizing Non-Covalent Interactions, Binzhou Lin

Theses and Dissertations

This thesis describes the development of molecular machines as analytical instruments to measure the strength of non-covalent interactions and to test hypotheses of reaction mechanisms. The first chapter is a review of molecular rotors. The second chapter presents the development of a molecular rotor that can be controlled using electrostatic interactions. Previous examples have used electrostatic interactions to slow down or inhibit molecular motion. Our system is activated and sped up using electrostatic interactions. The third and fourth chapters describe the application of molecular rotors designed to experimentally measure the strength and stability trends for non-covalent pnictogen and chalcogen interactions. …

New Techniques For Studying Ocean Eddy Dynamics In The Arabian Sea And The Gulf Of Mexico, Paul Andrew Ernst

Theses and Dissertations

Coherent vortices known as eddies are ubiquitous throughout the world’s oceans and are responsible for modulating regional climates via the transfer of heat, momentum, and salt. Two regions where eddies are particularly prominent are the Arabian Sea (AS) and the Gulf of Mexico (GoM). This dissertation advances the study of eddies in these regions by developing and employing new techniques in conjunction with remotely sensed data and high-resolution ocean models. These models focus particularly on lesser-studied eddy components at the subsurface and the submesoscale (below 25 km in radius).

The first component of this dissertation focuses on the Lakshadweep High …

On Parallelization Of Graph Algorithms, Performance Modelling And Autonomous 3d Printable Object Synthesis, Shams-Ul-Haq Syed

Theses and Dissertations

The degree of hardware level parallelism offered by today’s GPU architecture makes it ideal for problem domains with massive inherent parallelism potential, fields such as computer vision, image processing, graph theory and graph computations. We have identified three problem areas for purpose of this research dissertation, under the umbrella of performance improvement by harnessing the power of GPUs for novel applications. The first area is concerned with k-vertex connectivity in graph theory, the second area deals performance evaluation using extended roofline models for GPU parallel applications and finally the third problem area is related to synthesis 3D printable objects from …

Scene Text Detection And Recognition Via Discriminative Representation, Liang Zhao

Theses and Dissertations

Scene texts refer to arbitrary text presented in an image captured by a camera in the real world. The tasks of scene text detection and recognition from complex images play a crucial role in computer vision, with potential applications in scene understanding, information retrieval, robotics, autonomous driving, etc. Despite the notable progress made by existing deep-learning methods, achieving accurate text detection and recognition remains challenging for robust real-world applications. The challenges in scene text detection and recognition stem from: 1) diverse text shapes, fonts, colors, styles, layouts, etc.; 2) countless combinations of characters with unfixed attributes for complete detection, coupled …

Automated Data-Flow Optimization For Digital Signal Processors, Madushan Thilina Abeysinghe

Theses and Dissertations

Digital signal processors (DSP), which are characterized by statically-scheduled Very-Long Instruction Word architectures and software-defined scratchpad memory, are currently the go-to processor type for low-power embedded vision systems, as exemplified by the DSP processors integrated into systems-on-chips from NVIDIA, Samsung, Qualcomm, Apple, and Texas Instruments. DSPs achieve performance by statically scheduling workloads, both in terms of data movement and instructions. We developed a method for scheduling buffer transactions across a data flow graph using data-driven performance models, yielding a 25% average reduction in execution time and a reduction of up to 85% DRAM utilization for randomly-generated data flow graphs. We …

One-Dimensional Cores Of Peano Continua, Shogo Starr

Theses and Dissertations

In this paper, we define a berbering of a Peano continuum, which is the space one gets by attaching a null sequence of circles to a countable dense subset of a Peano continuum. The topology of a berbering is one that recognizes which circles are small, in the sense that any neighborhood of the original Peano continuum will entirely contain all but finitely many of the attached circles. We then show that given a berbering of any simply connected, locally simply connected Peano continuum $Y$, there exists a one-dimensional Peano continuum $X$ and a map from $X$ to the berbering …