Physical Sciences and Mathematics Commons^™

Deep Active Genetic Learning With Evidential Uncertainty For Agriculture Crops And Lake Water Quality Assessment, Oguz M. Aranay

Legacy Theses & Dissertations (2009 - 2024)

Despite significant advancements in the field of machine learning, there are two issues that still require further exploration. First, how to learn from a small dataset; and second, how to select appropriate features from the data. Although there exist many techniques to address these issues, choosing a combination of the techniques from these two groups is challenging, and worth investigating. To address these concerns, this thesis presents a learning framework that is based on a deep learning model utilizing active learning (with evidential uncertainty as a basis for acquisition function) for the first issue and a genetic algorithm for the …

Frameworks For Secure Collaborative And Concurrent Editing, Shashank Arora

Legacy Theses & Dissertations (2009 - 2024)

Cloud-based online document editing services, such as Google Docs and Office 365, provide an inexpensive and efficient means of managing documents. However, storing data on the cloud also raises certain security and privacy concerns, especially when the data is of confidential and sensitive nature. Storing data on third-party servers can potentially be compromising as it gives an opportunity to the third-party cloud service providers to turn semi-honest and become curious about user data. User data stored on third-party servers is also prone to attacks like virtual machine-based side-channel along with natural language processing and machine learning-based content search and retrieval …

An Ims-Ms/Md Workflow For Determining Higher Order Structure And Dynamics Of Nucleic Acids, Rebecca D'Esposito

Legacy Theses & Dissertations (2009 - 2024)

Ion mobility spectrometry - mass spectrometry (IMS-MS) has potential for the investigation of structure and dynamics in large biopolymers, which will come to full fruition only with a firmer understanding of how to interpret the experimental data. Numerous studies have employed elements of nucleic acid (NA) secondary structure, such as duplexes and hairpins, to explore the relationships between structure, experimental conditions, and actual observations. When combined with molecular dynamics simulations (MDS), IMS-MS can be effectively employed to perform structural elucidation of biomolecules that are not readily amenable to established techniques employed for structural analysis.

Human-Wildlife Coexistence With Coyotes In Los Angeles County, Ca And Cook County, Il, Alan Eapen

Legacy Theses & Dissertations (2009 - 2024)

The expansion of coyotes (Canis latrans) into urban and suburban areas where human concentration is high has led to human-coyote conflicts often leading to increased management actions against coyotes. Recently in Los Angeles and Cook Counties, municipalities have developed coyote management plans to foster human-wildlife coexistence in an effort to reduce conflict, an emerging concept that promotes the cohabitation of humans and animals in shared landscapes. The thesis investigates coyote management plans and policies concerning human-coyote interactions in Los Angeles and Cook Counties to address human-wildlife. Using a case study analysis of Los Angeles and Cook Counties, this study analyzed …

Examining The Impact Of Stratospheric Vortex Variability On Us Surface Temperature, Elena M. Fernandez

Legacy Theses & Dissertations (2009 - 2024)

Variability in high-latitude stratospheric flow, including major sudden stratospheric warming (SSW) or strong polar vortex events, can modulate tropospheric circulation and have meteorological implications on surface temperature and weather. The largest tropospheric impacts from Northern Hemisphere stratospheric variability – and the regions of focus in the literature – are observed in the North Atlantic, Europe and Asia. Impacts on North America occur but have received less research attention. The goal of this thesis research is to identify and quantify the relationship between high-latitude stratospheric variability and North American wintertime temperature.Using the ERA-Interim and ERA-5 reanalysis for the period of 1999/2000–2018/19, …

A Comparative Analysis Of The Impact Of Low-Level Jets And Atmospheric Rivers In The Central U.S, Nabindra Gyawali

Legacy Theses & Dissertations (2009 - 2024)

Atmospheric Rivers (AR) are the primary source of poleward moisture transport globally. In theUnited States, much recent attention has been placed on Pacific Coast ARs which occur between December and March. However, the central US is also substantially impacted by warm-season AR. An open question is whether these warm-season ARs are synonymous with Great Plains low-level jets (LLJs) that have long served as a focal point of mesoscale atmospheric research. In this study, we perform a comparative analysis of ARs and LLJs in the central U.S using ECMWF’s climate-quality reanalysis of the 20th century, CERA-20C. The aim is to identify …

Changes In Large-Scale Extreme Precipitation In Over Taiwan And The Northeast United States : Past And Future, Lexi Henny

Legacy Theses & Dissertations (2009 - 2024)

Large-scale extreme precipitation over (1) the mid-Atlantic and Northeast United States and (2) Taiwan is attributed to weather types such as atmospheric river (AR), TC, and extreme integrated vapor transport (IVT). Statistically significant increases in season-total EP day precipitation are seen at many GHCN stations in winter, summer, and fall in the mid-Atlantic and Northeast region, and at certain high-elevation grid points in Taiwan Mei-yu season. During the cold season of winter and spring, the U.S.-based changes come from AR-associated EP days and are associated with strengthened southwesterly winds and IVT either within EP days, in the season mean, or …

Multiple Imputation In High-Dimensional Data With Variable Selection, Qiushuang Li

Legacy Theses & Dissertations (2009 - 2024)

This dissertation focuses on the development of multiple imputation models and algorithms for high-dimensional data with variable selection structures. Leveraging on the multivariate linear mixed-effects model with missing responses for clustered data, we incorporate the variable selection routines using spike-and-slab priors within the Bayesian variable selection framework. Specific choice of these priors allow us to "force'' variables of importance (e.g. design variables or variables known to play role in missingness mechanism) into the imputation models. Our ultimate goal is to improve computational speed by removing unnecessary variables. Markov chain Monte Carlo techniques have been designed to sample from the implied …

Amyloid Fibril Formation And Polymorphism : A Critical Role Of Sulfur-Containing Amino Acid Residues, Tatiana Quiñones-Ruiz

Legacy Theses & Dissertations (2009 - 2024)

Protein aggregation that results in the formation of amyloid fibrils has been linked to many neurodegenerative disorders, including Alzheimer’s disease and Parkinson’s disease. The sulfur atoms in methionine (Met) and cysteine (Cys) residues of proteins can be readily oxidized, significantly affecting their properties. Oxidation of sulfur-containing amino acids has recently been shown to affect protein fibrillation. This work presents novel findings on Cys and Met redox reactions that are related to the formation of amyloid fibrils and on the polymorphism of a model fibrillogenic protein, hen egg white lysozyme (HEWL). Biophysical techniques including Raman spectroscopy, atomic force microscopy, electron paramagnetic …

Presto : Fast And Effective Group Closeness Maximization, Baibhav L. Rajbhandari

Legacy Theses & Dissertations (2009 - 2024)

Given a graph and an integer k, the goal of group closeness maximization is to find, among all possible sets of k vertices (called seed sets), a set that has the highest group closeness centrality. Existing techniques for this NP-hard problem strive to quickly find a seed set with a high, but not necessarily the highest centrality.

Calculational Methods In Conformal Field Theory, Thomas Andrew Vandermeulen

Legacy Theses & Dissertations (2009 - 2024)

We examine conformal field theories and their gaugings (orbifolds), first on the torus and then on surfaces of higher genus. Specific attention is paid to performing explicit calculations, to which end we introduce a number of specialized formalisms. Namely, the technology of flavored partition functions allows for a general treatment of orbifolds in theories with conserved currents, and Arakelov geometry facilitates the derivation of novel expressions for geometric quantities appearing commonly on higher genus worldsheets. Ultimately, we leverage the resulting expressions in the calculation of correlation functions via degeneration of Riemann surfaces.

Computational Methods For Propagation Of Optical Fields With The Angle-Impact Wigner Function, Jeremy Wittkopp

Legacy Theses & Dissertations (2009 - 2024)

In designing an optical setup for an experiment, one usually turns to simulations first in order to model the propagation of light through the proposed system. This way, the experimenter can determine if the system is operating as intended. In order for these simulations to be useful, they need to properly describe the propagation of light. In order to simplify calculations, most contemporary software makes assumptions on the nature of the light being propagated. Specifically, simulations typically consider optical fields that are beam-like (i.e., most of the rays comprising the field deviate only slightly in angle from the beam's primary …

Stability And Differential Privacy Of Stochastic Gradient Methods, Zhenhuan Yang

Legacy Theses & Dissertations (2009 - 2024)

Recently there are a considerable amount of work devoted to the study of the algorithmic stability as well as differential privacy (DP) for stochastic gradient methods (SGM). However, most of the existing work focus on the empirical risk minimization (ERM) and the population risk minimization problems. In this paper, we study two types of optimization problems that enjoy wide applications in modern machine learning, namely the minimax problem and the pairwise learning problem.

Probing The Stability And Solution Processability Of Metal Chalcogenide Semiconducting Materials, Mengwen Yan

Legacy Theses & Dissertations (2009 - 2024)

Metal chalcogenide (MCh) semiconductors have long research history due to their earth abundance, easy synthesizability, and accessible band gap tunability. People have realized their functionality as p-type semiconductors that provide good hole mobility and conductivity within the materials. However, MChs have easier reaction pathway with gas molecules (H2O and O2) compared to other semiconductors such as metal oxides. Also, poor solution processability of all MChs make low-cost thin film fabrication methods hard to achieve. In this thesis, we target to understand both the chemical stability and solution processability of MCh materials. ZnSe based inorganic organic layered hybrids (LHs) were exfoliated …

Raman Spectroscopy Of Aged Waterlogged Bones: Protocol Optimization, Bailey Hoplight

Chemistry

Current methods used for determining the age buried bones for forensic purposes are heavily affected by environmental conditions, can be susceptible to scientific bias and the interpretation can vary from investigator to investigator. Raman Spectroscopy can potentially be used as an automatic method to analyze the composition of bone eliminating the human factor. To test this hypothesis, we will use Raman spectroscopy to potentially identify the burial time of bones. The change in the chemical composition of the pig bones was probed using a Raman spectrometer with a 785nm excitation wavelength. Other parameters of the Raman instrument were optimized to …

The Analysis Of Relationships Between Lightning Strikes And Particulate Matter 2.5 Utilizing Statistical And Numerical Modeling Methods Applied To The 2020 Wildfire Season, Megan Schiede

Atmospheric & Environmental Sciences

Throughout 2020, ambient air pollution was reduced as a result from limiting anthropogenic activities to mitigate the spread of COVID-19. Not all air pollution is created the same as measurements of particulate matter 2.5 (PM2.5) were generally unaffected by this reduction due to PM2.5s source from wildfires. Despite influences from COVID-19 lockdowns, rises in PM2.5 concentrations can be attributed to the anomalously active wildfire season of 2020. As climate change progresses, these extraordinarily active seasons can be classified as the “new normal”; thus, comprehension of such events are vital. Given the ability of lightning to naturally cause wildfires, there exists …

Observations And Analysis Of The 7 October 2020 Derecho, Maxim Couillard

Atmospheric & Environmental Sciences

A serial derecho impacted Upstate New York and western Massachusetts on the afternoon of 7 October 2020, resulting in 120 damaging wind reports alonga 515-km swath. The magnitude and extent of this event was unexpected and not well forecast, motivating a closer examination. This study seeks to better understand the conditions prior to and during the derecho to diagnose what led to such a damaging wind event. The synoptic forcing for this event was an upper- level jet with a broad, negatively tilted trough over the northeast U.S. and southeast Canada. The upper-level pattern provided forcing for a 990-hPa surface …

Space Weather And Criminal Violence : A Longitudinal Analysis Of Major Us Urban Areas, Richard Mcmillan

Legacy Theses & Dissertations (2009 - 2024)

The study of criminology began with the investigation of the physical features of humans committing crimes such as homicide. The study of space weather includes the investigation of the physiological manifestations of various forms of space weather on terrestrial-bound humans. These two fields of inquiry are joined in this investigation of the association of space weather with the incidence of criminal violence in the United States. This study of the possibility that a physical force emanating from the Sun or from outer space can affect human behavior in the form of criminal violence is part of a long search for …

Hierarchical Modelling Of Raman Spectroscopic Data Demonstrates The Potential For Manufacturer And Caliber Differentiation Of Smokeless Powders, Nathan Russell Bartlett

Legacy Theses & Dissertations (2009 - 2024)

With the rise of gun-based crime and the advent of lead-free ammunition reducing the effectiveness of previously used methods, the analysis of smokeless powders to better understand organic gunshot residues is now more important than ever. One method that has shown potential is Raman spectroscopy, a relatively quick, nondestructive, and non-contact analysis method. To investigate whether Raman spectroscopy could be used to characterize individual powders, samples of smokeless powders from three manufacturers, Remington, Federal, and Winchester, each with two 9mm caliber and two .38 caliber ammo types, were dissolved, and their Raman spectra were collected. A hierarchical modelling system was …

Permaculture As An Ecopedagogy Curriculum And Alternative Theory Of Development : An Exploration Of The Ecological Consciousness Of Rural Western Kenyan Farmers Using Photo-Voice With A Farmer Field School During Covid-19, David Yisrael Epstein

Legacy Theses & Dissertations (2009 - 2024)

This dissertation explores two main questions of farmers in Western Kenya – how they see farming as a creative process and their role in that – and how change is made possible in their community. I explore stories from farmers themselves, using photographs they have taken and stories they tell about those photographs, which answer these two questions. In doing so, I attempt to understand the degree to which collective action has taken root in a community exposed to a permaculture based curriculum within a farmer field school. It also importantly seeks to understand what happens when such a curriculum …

Exposing Gan-Generated Faces Using Deep Neural Network, Hui Guo

Legacy Theses & Dissertations (2009 - 2024)

Generative adversarial network (GAN) generated high-realistic human faces are visually challenging to discern from real ones. They have been used as profile images for fake social media accounts, which leads to high negative social impacts.In this work, we explore a universal physiological cue of the eye, namely the pupil shape consistency, to identify GAN-generated faces reliably. We show that GAN-generated faces can be exposed via irregular pupil shapes. This phenomenon is caused by the lack of physiological constraints in the GAN models. We demonstrate that such artifacts exist widely in high-quality GAN-generated faces. We design an automatic method to segment …

Long-Term Trends And Variability Of Pm₂.₅ Mass, Chemical Components, And Sources Over Nys, Hesham Abdelmonem Hassan

Legacy Theses & Dissertations (2009 - 2024)

Eighteen years of ambient PM2.5 mass and major components are used in this study to harmonize the carbonaceous fraction of PM2.5 data, and to examine long-term trends and seasonal variability changes. The major decreasing trend of PM2.5 mass and major inorganic ions reflects the ambient concentration response to clean air regulations and policies.Although elemental carbon shows a decreasing trend in agreement with most other PM2.5 components, organic carbon shows a different behavior of no-trend to slightly increasing trend in some sites. Organic carbon behavior indicates the growing role of non-combustion sources at the expense of decreasing combustion sources. Seasonal trends …

Stubbornly Merging Discrete Vector Fields, Douglas W. Lenseth

Legacy Theses & Dissertations (2009 - 2024)

We address the basic question in discrete Morse theory of combining discrete gradientfields that are partially defined on subsets of the given complex. This is a well-posed question when the discrete gradient field V is generated using a fixed algorithm which has a local nature. One example is ProcessLowerStars, a widely used algorithm for computing persistent homology associated to a grayscale image in 2D or 3D. While the algorithm for V may be inherently local, being computed within stars of vertices and so embarrassingly parallelizable, in practical use it is natural to want to distribute the computation over patches Pi, …

An Entropic Approach To Dynamics, Pedro Henrique Moreira Pessoa

Legacy Theses & Dissertations (2009 - 2024)

The present thesis deals with different aspects of probability, the method of maximum entropy, information geometry, and dynamical systems and their applications to statistical physics and complex systems. Those topics come together in the framework for dynamics termed entropic dynamics. Work on the topic was initiated by my PhD advisor Professor Ariel Caticha and many colleagues at University at Albany as a method for explaining the dynamical processes in quantum mechanics from first principles of probability theory.

Signal Yields And Detector Modeling In Xenon Time Projection Chambers, And Results Of An Effective Field Theory Dark Matter Search Using Lux Data, Gregory Ransford Carl Rischbieter

Legacy Theses & Dissertations (2009 - 2024)

The nature of dark matter continues to be one of the biggest remaining mysteries in physics. Astrophysical measurements indicate that dark matter makes up more than a quarter of the Universe's total energy density, and it is well-motivated that dark matter is comprised of Weakly Interacting Massive Particles (WIMPs). Direct detection techniques utilizing liquid and gaseous noble elements have become the primary method of probing the potential non-gravitational interactions between WIMPs and Standard Model matter, with the leading technology being the dual-phase Time Projection Chamber (TPC). The Large Underground Xenon (LUX) and its second-generation successor, LUX-ZEPLIN (LZ), are two xenon …

The Mountain Gap Wind Of The Isthmus Of Tehuantepec And Its Impact On Tropical Cyclogenesis In The Eastern Pacific, Adolfo Lugo Rios

Atmospheric and Environmental Sciences Theses and Dissertations

In this dissertation we research two aspects of the mountain gap wind northerlies of the Isthmus of Tehuantepec in Southern Mexico. The causes of extreme gap wind northerlies, emphasizing the causes by tropical weather systems, like easterly waves troughs or hurricanes, and the impacts of an extreme gap wind events on tropical cyclogenesis.

Through the use of an index to diagnose the mountain gap wind strength, its evolution through the year and the mechanisms leading to extreme mountain gap winds was highlighted. The index revealed that the gap wind is strongly related to high surface pressure anomalies over the GoM …

Fluorescent Labelling Of Ribosomal Rna In Live Mammalian Cells, Amir Agharezaee

Legacy Theses & Dissertations (2009 - 2024)

This thesis describes a bio-orthogonal chemistry-based approach for new methods of fluorescence labelling of RNA in live HeLa cells via an IEDDA reaction of TCO with tetrazine. The key step of the approach is synthesis of two nucleobase analogues which were modified with TCO and a masked 5′-phosphate group using aryl phosphonamidite chemistry. Synthesis of tubercidin and uridine analogues were considered in this investigation for incorporation into RNA in live HeLa cells. Synthetic approach was successful at installing the handle containing TCO to tubercidin and uridine to make compounds 12 and 21 respectively (the synthesis of both of which are …

Dynamic Impacts Of Hadley Circulation On Saharan Desert Warming Amplification, Alejandro Manuel Ayala

Legacy Theses & Dissertations (2009 - 2024)

Changes in temperature due to climate change are not spatially uniform, and deserts and other drylands, which are greatly underrepresented in climate studies, are warming at a much faster rate than much of the globe with increasing concentrations of greenhouse gases. This strong warming amplification over deserts, termed Desert Amplification (DA), is most pronounced over the world’s largest and driest Sahara Desert and the Arabian Peninsula. The Sahara and Arabian deserts are formed in the subtropical subsiding branch of the Hadley Circulation (HC) and so the changes in large-scale subsidence associated with adiabatic heating could impact the DA dynamically. While …

X Ray Phase And Coherent Scatter Imaging Measurements, Pikting Cheung

Legacy Theses & Dissertations (2009 - 2024)

Traditional x-ray images are good for distinguishing between materials with very different absorption coefficients. However, detecting breast cancer using conventional x-ray imaging is hard because cancer tissue has similar absorption coefficients as normal tissue. Annually 43,600 women will die from breast cancer. The odds of survival are better if the cancer is diagnosed at an early stage, but mammography tends to have high rates of missed cancers. Mammography also tends to have a high false-positive rate, which leads to unnecessary biopsies. Two techniques with potential to add additional information to breast imaging have been investigated. X-ray phase imaging has been …

Spatiotemporal Variability Of Snow Depth On Arctic Sea Ice In Cmip6 Models And Satellite Data, Shengzhe Chen

Legacy Theses & Dissertations (2009 - 2024)

The Arctic has experienced rapid environmental changes in recent decades. Motivated by the important role of snow over sea ice in influencing Arctic climate variability and change, this dissertation aims to improve our knowledge of spatiotemporal variability of snow depth and processes over Arctic sea ice in the observations and model simulations.First, we evaluate snow depth over Arctic sea ice during 1993–2014 simulated by the models from the Coupled Model Intercomparison Project phase 6 (CMIP6) against recent satellite retrievals. The CMIP6 models capture some aspects of the observed snow depth climatology and variability. The observed variability lies in the middle …