Physical Sciences and Mathematics Commons^™

Generalized Differentiable Neural Architecture Search With Performance And Stability Improvements, Emily J. Herron

Doctoral Dissertations

This work introduces improvements to the stability and generalizability of Cyclic DARTS (CDARTS). CDARTS is a Differentiable Architecture Search (DARTS)-based approach to neural architecture search (NAS) that uses a cyclic feedback mechanism to train search and evaluation networks concurrently, thereby optimizing the search process by enforcing that the networks produce similar outputs. However, the dissimilarity between the loss functions used by the evaluation networks during the search and retraining phases results in a search-phase evaluation network, a sub-optimal proxy for the final evaluation network utilized during retraining. ICDARTS, a revised algorithm that reformulates the search phase loss functions to ensure …

Material Formulation And Process Optimization Towards Fabricating Robust 3d Printed Structures, Austin Riggins

Doctoral Dissertations

This dissertation focuses on understanding and addressing the fundamental physicochemical phenomena that lead to weak interfaces and structural warpage in material extrusion 3D printing. Polymeric feedstocks used for this manufacturing technique were manipulated through the incorporation of additives that alter the dynamics of the matrix during and after printing. In Chapter II, adhesion between layers of structures printed from PEEK was strengthened through a combination of low-molecular weight additive incorporation and post-printing thermal annealing. Chapter III reports a method for decreasing the irreversible thermal strain of structures printed from poly(lactic acid) by introducing nanographene and photoinitiator additives into the feedstock …

Static And Dynamic State Estimation Applications In Power Systems Protection And Control Engineering, Ibukunoluwa Olayemi Korede

Doctoral Dissertations

The developed methodologies are proposed to serve as support for control centers and fault analysis engineers. These approaches provide a dependable and effective means of pinpointing and resolving faults, which ultimately enhances power grid reliability. The algorithm uses the Least Absolute Value (LAV) method to estimate the augmented states of the PCB, enabling supervisory monitoring of the system. In addition, the application of statistical analysis based on projection statistics of the system Jacobian as a virtual sensor to detect faults on transmission lines. This approach is particularly valuable for detecting anomalies in transmission line data, such as bad data or …

Atomic-Level Mechanisms Of Fast Relaxation In Metallic Glasses, Leo W. Zella

Doctoral Dissertations

Glasses are ubiquitous in daily life and have unique properties which are a consequence of the underlying disordered structure. By understanding the fundamental processes that govern these properties, we can modify glasses for desired applications. Key to understanding the structure-dynamics relationship in glasses is the variety of relaxation processes that exist below the glass transition temperature. Though these relaxations are well characterized with macroscopic experimental techniques, the microscopic nature of these relaxations is difficult to elucidate with experimental tools due to the requirements of timescale and spatial resolution. There remain many questions regarding the microscopic nature of relaxation in glass …

Exploring Soil Microbial Dynamics In Southern Appalachian Forests: A Systems Biology Approach To Prescribed Fire Impacts, Saad Abd Ar Rafie

Doctoral Dissertations

Prescribed fires in Southern Appalachian forests are vital in ecosystem management and wildfire risk mitigation. However, understanding the intricate dynamics between these fires, soil microbial communities, and overall ecosystem health remains challenging. This dissertation addresses this knowledge gap by exploring selected aspects of this complex relationship across three interconnected chapters.

The first chapter investigates the immediate effects of prescribed fires on soil microbial communities. It reveals subtle shifts in porewater chemistry and significant increases in microbial species richness. These findings offer valuable insights into the interplay between soil properties and microbial responses during the early stages following a prescribed fire. …

A Measurement Of Neutron Polarization And Transmission For The Nedm@Sns Experiment, Kavish Imam

Doctoral Dissertations

The D.O.E Nuclear Science Advisory Committee Long Range Plan has called for experimental programs to explore fundamental symmetry violations and their implications in nuclear, particle and cosmological physics. The neutron electric dipole moment experiment at the Spallation Neutron Source (nEDM@SNS) aims to search for new physics in the Time-reversal (T) and Charge-Parity (CP) symmetry violating sector by setting a new limit on the nEDM down to a few x 10^-28 e·cm using a novel cryogenic technique, which combines the unique properties of polarized Ultracold Neutrons (UCN), polarized ³He, and superfluid ⁴He. The experiment will employ a cryogenic …

Towards Safer Code Reuse: Investigating And Mitigating Security Vulnerabilities And License Violations In Copy-Based Reuse Scenarios, David Reid

Doctoral Dissertations

Background: A key benefit of open source software is the ability to copy code to reuse in other projects. Code reuse provides benefits such as faster development time, lower cost, and improved quality. There are several ways to reuse open source software in new projects including copy-based reuse, library reuse, and the use of package managers. This work specifically looks at copy-based code reuse.

Motivation: Code reuse has many benefits, but also has inherent risks, including security and legal risks. The reused code may contain security vulnerabilities, license violations, or other issues. Security vulnerabilities may persist in projects that copy …

Characterizing Silicate Materials Via Raman Spectroscopy And Machine Learning: Implications For Novel Approaches To Studying Melt Dynamics, Blake O. Ladouceur

Doctoral Dissertations

Silicate melt characteristics impose dramatic influence over igneous processes that operate, or have operated on, differentiated bodies: such as the Earth and Mars. Current understanding of these melt properties, such as composition, primarily comes from investigations on their volcanic byproducts. Therefore, it is imperative to innovate on modalities capable of constraining melt information in environments where a reliance on laboratory methods is severed. Recent investigations have turned to Raman Spectroscopy and amorphous volcanics as a suitable pairing for exploring these ideas. Silicate glasses are a proxy for igneous melts; and Raman spectroscopy is a robust analytical technique capable of operating …

Integration Of Raman Spectroscopy And Python-Based Data Analysis For Advancing Neurobiological Research, Natalie E. Dunn

Doctoral Dissertations

The field of Raman spectroscopy continues to expand into biological applications due to its usefulness as a non-invasive technique that can be utilized qualitatively and quantitatively. However, the inherent weakness of Raman scattering leads to the need for each collected spectra to undergo a preprocessing step to remove noise, background drift, and cosmic rays. Biological research in particular needs large datasets due to the increased variability in samples. As datasets grow, the need to perform preprocessing on each individual spectra becomes daunting. Often, these steps are done by hand with the help of specialized software programs. Preprocessing can be accelerated …

Towards Expressive And Versatile Visualization-As-A-Service (Vaas), Tanner C. Hobson

Doctoral Dissertations

The rapid growth of data in scientific visualization has posed significant challenges to the scalability and availability of interactive visualization tools. These challenges can be largely attributed to the limitations of traditional monolithic applications in handling large datasets and accommodating multiple users or devices. To address these issues, the Visualization-as-a-Service (VaaS) architecture has emerged as a promising solution. VaaS leverages cloud-based visualization capabilities to provide on-demand and cost-effective interactive visualization. Existing VaaS has been simplistic by design with focuses on task-parallelism with single-user-per-device tasks for predetermined visualizations. This dissertation aims to extend the capabilities of VaaS by exploring data-parallel visualization …

Short Range Correlation Measurements In The Quasielastic Region With An 11 Gev Beam, Casey Morean

Doctoral Dissertations

Electron scattering is a significant means of studying internal high momentum
nucleon and quark distributions in nuclei. Thomas Jefferson National Accelerator
Facility (JLab) with its 11GeV beam is capable of studying high momentum nucleons
with unmatched precision. The role of short range nucleon configurations and
quark distributions is significant for understanding the dynamics of nuclei and their
underlying components. Scattering cross section measurements in the kinematic
regime x > 1, where the free nucleon is forbidden, are sensitive to high momentum
nucleons, which are believed to come from short range correlations (SRCs). SRCs are
strongly interacting, high momentum nucleons with a …

Exact Models, Heuristics, And Supervised Learning Approaches For Vehicle Routing Problems, Zefeng Lyu

Doctoral Dissertations

This dissertation presents contributions to the field of vehicle routing problems by utilizing exact methods, heuristic approaches, and the integration of machine learning with traditional algorithms. The research is organized into three main chapters, each dedicated to a specific routing problem and a unique methodology. The first chapter addresses the Pickup and Delivery Problem with Transshipments and Time Windows, a variant that permits product transfers between vehicles to enhance logistics flexibility and reduce costs. To solve this problem, we propose an efficient mixed-integer linear programming model that has been shown to outperform existing ones. The second chapter discusses a practical …

Generative Adversarial Game With Tailored Quantum Feature Maps For Enhanced Classification, Anais Sandra Nguemto Guiawa

Doctoral Dissertations

In the burgeoning field of quantum machine learning, the fusion of quantum computing and machine learning methodologies has sparked immense interest, particularly with the emergence of noisy intermediate-scale quantum (NISQ) devices. These devices hold the promise of achieving quantum advantage, but they grapple with limitations like constrained qubit counts, limited connectivity, operational noise, and a restricted set of operations. These challenges necessitate a strategic and deliberate approach to crafting effective quantum machine learning algorithms.

This dissertation revolves around an exploration of these challenges, presenting innovative strategies that tailor quantum algorithms and processes to seamlessly integrate with commercial quantum platforms. A …

Migmatite Formation, Geochronometer Petrogeneisis, And Rare Earth Element Mineralization In The Adirondack Mountains, Ny, Kaitlyn Suarez

Doctoral Dissertations

The Adirondack Mountains in upstate New York contain exposures of complex partially melted rocks, in addition to iron oxide-apatite (IOA) deposits with variable rare earth element (REE) concentrations. Previous workers have suggested that melting occurred during the ca. 1150 Ma Shawinigan and the ca. 1050 Ma Ottawan orogenies. However, there are challenges in determining the timing of melting and the number of partial melt events. Further, tectonic models must be developed to describe the petrogenesis of IOA and REE mineralization. Migmatites are present along Rt. 4/22 near Whitehall, NY. In chapter two, all layers of a single migmatitic rock were …

Polymer-Based Nanotherapeutics To Combat Difficult-To-Treat Bacterial Infections, Jessa Marie V. Makabenta

Doctoral Dissertations

The continuous emergence and spread of antibiotic-resistant bacteria are a global health emergency, debilitating the capability to prevent and cure various infectious diseases that were once treatable. Antibiotic therapy is further rendered ineffective due to biofilm formation and the ability of bacteria to thrive and colonize inside mammalian cells. Given the diminishing efficacy of available antibiotics combined with the scarcity of new therapeutics entering the antibiotic pipeline, innovative treatment strategies are urgently in demand. Nanomaterial-based strategies offer ‘outside of the box’ approach for the treatment of antibiotic-resistant bacterial infections. Nanomaterials feature tunable physicochemical properties that can be carefully modified to …

Intersection Cohomology Of Rank One Local Systems For Arrangement Schubert Varieties, Shuo Lin

Doctoral Dissertations

In this thesis we study the intersection cohomology of arrangement Schubert varieties with coefficients in a rank one local system on a hyperplane arrangement complement. We prove that the intersection cohomology can be computed recursively in terms of certain polynomials, if a local system has only $\pm 1$ monodromies. In the case where the hyperplane arrangement is generic central or equivalently the associated matroid is uniform and the local system has only $\pm 1$ monodromies, we prove that the intersection cohomology is a combinatorial invariant. In particular when the hyperplane arrangement is associated to the uniform matroid of rank $n-1$ …

When To Hold And When To Fold: Studies On The Topology Of Origami And Linkages, Mary Elizabeth Lee

Doctoral Dissertations

Linkages and mechanisms are pervasive in physics and engineering as models for a
variety of structures and systems, from jamming to biomechanics. With the increase
in physical realizations of discrete shape-changing materials, such as metamaterials,
programmable materials, and self-actuating structures, an increased understanding
of mechanisms and how they can be designed is crucial. At a basic level, linkages
or mechanisms can be understood to be rigid bars connected at pivots around which
they can rotate freely. We will have a particular focus on origami-like materials, an
extension to linkages with the added constraint of faces. Self-actuated versions typ-
ically start …

Reactive Chemistries For Protein Labeling, Degradation, And Stimuli Responsive Delivery, Myrat Kurbanov

Doctoral Dissertations

Reactive chemistries for protein chemical modification play an instrumental role in chemical biology, proteomics, and therapeutics. Depending on the application, the selectivity of these modifications can range from precise modification of an amino acid sequence by genetic manipulation of protein expression machinery to a stochastic modification of lysine residues on the protein surface. Ligand-Directed (LD) chemistry is one of the few methods for targeted modification of endogenous proteins without genetic engineering. However, current LD strategies are limited by stringent amino acid selectivity. To bridge this gap, this thesis focuses on the development of highly reactive LD Triggerable Michael Acceptors (LD-TMAcs) …

Towards Robust Long-Form Text Generation Systems, Kalpesh Krishna

Doctoral Dissertations

Text generation is an important emerging AI technology that has seen significant research advances in recent years. Due to its closeness to how humans communicate, mastering text generation technology can unlock several important applications such as intelligent chat-bots, creative writing assistance, or newer applications like task-agnostic few-shot learning. Most recently, the rapid scaling of large language models (LLMs) has resulted in systems like ChatGPT, capable of generating fluent, coherent and human-like text. However, despite their remarkable capabilities, LLMs still suffer from several limitations, particularly when generating long-form text. In particular, (1) long-form generated text is filled with factual inconsistencies to …

Human-Centered Technologies For Inclusive Collection And Analysis Of Public-Generated Data, Mahmood Jasim

Doctoral Dissertations

The meteoric rise in the popularity of public engagement platforms such as social media, customer review websites, and public input solicitation efforts strives for establishing an inclusive environment for the public to share their thoughts, ideas, opinions, and experiences. Many decisions made at a personal, local, or national scale are often fueled by data generated by the public. As such, inclusive collection, analysis, sensemaking, and utilization of pubic-generated data are crucial to support the exercise of successful decision-making processes. However, people often struggle to engage, participate, and share their opinions due to inaccessibility, the rigidity of traditional public engagement methods, …

Quantifying And Enhancing The Security Of Federated Learning, Virat Vishnu Shejwalkar

Doctoral Dissertations

Federated learning is an emerging distributed learning paradigm that allows multiple users to collaboratively train a joint machine learning model without having to share their private data with any third party. Due to many of its attractive properties, federated learning has received significant attention from academia as well as industry and now powers major applications, e.g., Google's Gboard and Assistant, Apple's Siri, Owkin's health diagnostics, etc. However, federated learning is yet to see widespread adoption due to a number of challenges. One such challenge is its susceptibility to poisoning by malicious users who aim to manipulate the joint machine learning …

Sources And Controls Of Carbon Dioxide In Inland Waters At Watershed, Regional, And Continental Scales, Brian Saccardi

Doctoral Dissertations

Inland waters are significant sources of carbon dioxide to the atmosphere, and estimates of emissions are similar in magnitude to those of the carbon dioxide sequestered by the net terrestrial sink. Currently, methods of estimating carbon dioxide emissions are based on statistical approaches and often do not consider landscape attributes such as human development, agriculture, or the hydrologic connectivity of the stream network. The following research addresses these issues in chapter 1 by developing and validating a reactive transport model at the watershed scale, then in chapter 2 by applying the reactive transport model at the continental scale across US …

Search For Exotic Higgs Boson Decay To Multiple B-Quarks With The Atlas Detector At Lhc Using Machine Learning Methods, Yuan-Tang Chou

Doctoral Dissertations

The discovery of the Higgs boson has opened up new possibilities for investigating physics beyond the Standard Model (SM). New particles may interact with the SM through the Higgs boson, and deviations from SM predictions can indicate the presence of new physics. This dissertation focuses on the search for exotic Higgs decay, $H\rightarrow aa \rightarrow (b\bbar)(b\bbar)$ where a is a new scalar boson and focuses on the case where the Higgs boson is produced in association with a Z boson. The data were collected by the ATLAS detector at the Large Hadron Collider at center-of-mass energy $\sqrt{s} =~13~\TeV$ from 2015 …

Learning To See With Minimal Human Supervision, Zezhou Cheng

Doctoral Dissertations

Deep learning has significantly advanced computer vision in the past decade, paving the way for practical applications such as facial recognition and autonomous driving. However, current techniques depend heavily on human supervision, limiting their broader deployment. This dissertation tackles this problem by introducing algorithms and theories to minimize human supervision in three key areas: data, annotations, and neural network architectures, in the context of various visual understanding tasks such as object detection, image restoration, and 3D generation. First, we present self-supervised learning algorithms to handle in-the-wild images and videos that traditionally require time-consuming manual curation and labeling. We demonstrate that …

Foundations Of Node Representation Learning, Sudhanshu Chanpuriya

Doctoral Dissertations

Low-dimensional node representations, also called node embeddings, are a cornerstone in the modeling and analysis of complex networks. In recent years, advances in deep learning have spurred development of novel neural network-inspired methods for learning node representations which have largely surpassed classical 'spectral' embeddings in performance. Yet little work asks the central questions of this thesis: Why do these novel deep methods outperform their classical predecessors, and what are their limitations? We pursue several paths to answering these questions. To further our understanding of deep embedding methods, we explore their relationship with spectral methods, which are better understood, and show …

Thermodynamic Laws Of Billiards-Like Microscopic Heat Conduction Models, Ling-Chen Bu

Doctoral Dissertations

In this thesis, we study the mathematical model of one-dimensional microscopic heat conduction of gas particles, applying both both analytical and numerical approaches. The macroscopic law of heat conduction is the renowned Fourier’s law J = −k∇T, where J is the local heat flux density, T(x, t) is the temperature gradient, and k is the thermal conductivity coefficient that characterizes the material’s ability to conduct heat. Though Fouriers’s law has been discovered since 1822, the thorough understanding of its microscopic mechanisms remains challenging [3] (2000). We assume that the microscopic model of heat conduction is a hard ball system. The …

Positive Factorizations Via Planar Mapping Classes And Braids, Richard E. Buckman

Doctoral Dissertations

In this thesis we seek to better understand the planar mapping class group in
order to find factorizations of boundary multitwists, primarily to generate and study
symplectic Lefschetz pencils by lifting these factorizations. Traditionally this method
is applied to a disk or sphere with marked points, utilizing factorizations in the stan-
dard and spherical braid groups, whereas in our work we allow for multiple boundary components. Dehn twists along these boundaries give rise to exceptional sections of Lefschetz fibrations over the 2–sphere, equivalently, to Lefschetz pencils with base points. These methods are able to derive an array of known examples …

Experiments With Monopoles, Rings And Knots In Spinor Bose-Einstein Condensates, Alina A. Blinova

Doctoral Dissertations

Topological excitations are ubiquitous in nature, their charge being a naturally-quantized, conserved quantity that can exhibit particle-like behavior. Spinor Bose-Einstein condensates (BECs) are an exceptionally versatile system for the study and exploration of topological excitations. Between the spin-1 and spin-2 ⁸⁷Rb condensates there are seven possible broken-symmetry magnetic phases, with each one hosting unique opportunities for topological defects. We have created and observed several novel topological excitations in a spinor ⁸⁷Rb BEC. In this dissertation I present and discuss three principal experimental findings: (1) The discovery of an Alice ring, or a half-quantum vortex ring, emerging from a …

Probing The Physical Mechanisms Responsible For Brown Dwarf And Giant Planet Formation, Sarah Betti

Doctoral Dissertations

The disks that form around young stellar objects provide the essential material for their continued growth as well as the formation of planets, making them ideal laboratories to investigate the mechanisms and environments key for substellar and planetary formation. In this dissertation, I explore two main formation processes: the transportation of water necessary for giant planet formation, and the accretion and growth of young brown dwarfs. First, I study the water ice content in the circumstellar disk of AB Aurigae, a young Herbig Ae star. I detect and map icy grains on the disk surface using high contrast observations taken …

Nonparametric Derivative Estimation Using Penalized Splines: Theory And Application, Bright Antwi Boasiako

Doctoral Dissertations

This dissertation is in the field of Nonparametric Derivative Estimation using
Penalized Splines. It is conducted in two parts. In the first part, we study the L2
convergence rates of estimating derivatives of mean regression functions using penalized splines. In 1982, Stone provided the optimal rates of convergence for estimating derivatives of mean regression functions using nonparametric methods. Using these rates, Zhou et. al. in their 2000 paper showed that the MSE of derivative estimators based on regression splines approach zero at the optimal rate of convergence. Also, in 2019, Xiao showed that, under some general conditions, penalized spline estimators …