Physical Sciences and Mathematics Commons^™

Empowering Crisis Information Extraction Through Actionability Event Schemata And Domain-Adaptive Pre-Training, Yuhao Zhang, Siaw Ling Lo, Phyo Yi Win Myint

Research Collection School Of Computing and Information Systems

One of the persistent challenges in crisis detection is inferring actionable information to support emergency response. Existing methods focus on situational awareness but often lack actionable insights. This study proposes a holistic approach to implementing an actionability extraction system on social media, including requirement gathering, selection of machine learning tasks, data preparation, and integration with existing resources, providing guidance for governments, civil services, emergency workers, and researchers on supplementing existing channels with actionable information from social media. Our solution leverages an actionability schema and domain-adaptive pre-training, improving upon the state-of-the-art model by 5.5% and 10.1% in micro and macro F1 …

Defending Federated Recommender Systems Against Untargeted Attacks: A Contribution-Aware Robust Aggregation Scheme, Ruicheng Liang, Yuanchun Jiang, Feida Zhu, Ling Cheng, Huiwen Liu

Research Collection School Of Computing and Information Systems

Federated recommender systems (FedRSs) effectively tackle the tradeoff between recommendation accuracy and privacy preservation. However, recent studies have revealed severe vulnerabilities in FedRSs, particularly against untargeted attacks seeking to undermine their overall performance. Defense methods employed in traditional recommender systems are not applicable to FedRSs, and existing robust aggregation schemes for other federated learning-based applications have proven ineffective in FedRSs. Building on the observation that malicious clients contribute negatively to the training process, we design a novel contribution-aware robust aggregation scheme to defend FedRSs against untargeted attacks, named contribution-aware Bayesian knowledge distillation aggregation (ConDA), comprising two key components for the …

Geometric Dimensionality Reduction, Daniel Livschitz

CGU Theses & Dissertations

The emergence of AI models developed through computationally intensive training has resulted in a surge of research into dimensionality reduction techniques that spans across numerous mathematical disciplines. In this thesis we establish Geometric Dimensionality Reduction, a non-linear data compression technique that utilizes low dimensional manifolds embedded in dimensional spaces to form composite contraction-and-projection maps. Geometric Dimensionality Reduction is predominantly demonstrated through a novel algorithm entitled LGE (Livschitz-Gu-Eyunni) that utilizes Multicomplex rotation groups and polyspherical coordinates to define a single tuneable logarithmic map from ℝ 2푛 to ℝ 푛+1 with deterministic time complexity, geometric tunability, and semi-reversibility. Significant breakthroughs in the …

Predicting Superconducting Critical Temperature From Composition-Derived Features: A Transparent Linear And Regularized Regression Study, Md Ahiduzzaman

Data Science and Data Mining

We study prediction of superconducting critical temperature (Tc) from 81 composition-derived descriptors across 21,263 materials. To keep the analysis transparent and repro- ducible, we focus on linear models: Ordinary Least Squares (OLS), Ridge, Lasso, and Elastic Net (ENet). All models share a single evaluation protocol (5-fold cross-validation with standardized inputs) and are compared on RMSE, MAE, and R2. On this feature set, OLS attains the best cross-validated performance (RMSE = 17.6 K, MAE = 13.3 K , R2 = 0.735), with Lasso/ENet essentially tied next (RMSE ≈ 17.7 K , R2 ≈ 0.734); Ridge underperforms (RMSE = 18.9 K , …

Comparative Analysis Of Lasso, Ridge, And Elastic Net For Variable Selection In High-Dimensional Maize Data, Md Ahiduzzaman

Data Science and Data Mining

In high-dimensional genomic data analysis, traditional linear regression techniques often struggle due to the presence of a large number of predictor variables relative to observations. Penalized regression methods such as LASSO, Ridge, and Elastic Net have emerged as effective solutions by imposing regularization, which helps in managing multicollinearity and enhancing prediction accuracy. This study applies these techniques to the Maize dataset to model the time to male flowering, selecting relevant genetic markers as predictors. Our findings suggest that Elastic Net is particularly effective for high-dimensional data with correlated variables, achieving a balance between prediction accuracy and variable selection. The results …

Handwritten Digit Recognition Using Machine Learning Classifiers, Md Ahiduzzaman

Data Science and Data Mining

This project explores and compares the performance of various machine learning classifiers for handwritten digit recognition using the MNIST dataset. The classifiers include Logistic Regression, k-Nearest Neighbors, and Convolutional Neural Networks. Each classifier is evaluated based on accuracy, precision, recall, F1-score, and confusion matrix analysis.

Modification Of Nanoparticles For Designed Interfaces, Tony Lee Neely, Jr.

Theses and Dissertations

The work contained herein, is focused on the design, synthesis, and characterization of polymer nanocomposite interfaces and the property enhancement afforded from said interface design. Through the use of reversible addition fragmentation chain transfer (RAFT) polymerization for the grafting of polymer chains to silica nanoparticles, the surface of silica nanoparticles can be manipulated to tune the properties of the nanocomposite as a whole.

In the first part of this work, heterogeneity is introduced onto the surface of silica nanoparticles via a sequential RAFT polymerization to afford a bimodal brush system. A densely grafted, short brush population is polymerized from the …

A Computational Approach To The Game Of Cycles, Jocelyn Garcia, Mike Janssen, Eliza Kautz

Faculty Work Comprehensive List

In Mathematics for Human Flourishing, Francis E. Su introduced The Game of Cycles, a game played on a finite simple planar graph. In game play, opponents alternate adding direction to the edges of the graph with the goal of creating a cycle or making the last legal move. Recent work has sought to determine winning strategies on certain classes of graphs. We introduce a tabular representation of a game state and provide a computer program that determines which player has a winning strategy on any legal game board. The program builds a directed graph of all possible game states, utilizing …

Fabrication Of Functional Microparticles For Biological Analysis, Justin George

Electronic Theses & Dissertations (2024 - present)

To better advance our knowledge of cellular behavior, development of an organism, and disease mechanisms understanding cellular heterogeneity in the functional proteome is essential. Bulk measurements across a cell population often fail to capture the variations which exist between individual cells, however single cell proteomic technology has increasingly demonstrated the characterization of cellular heterogeneity. A single cell functional proteomics assay needs to be both high-throughput and multiplexed to effectively identify biomarkers and provide a quantitative correlation to better understand the interrelationships within complex protein networks. Janus particles have the ability to be engineered to have good compatibility and stability in …

Towards Human Explainable Digital Forensics: Generating Human Interpretable Evidence For Semantic Understanding In Manipulated Images And Text, Yuwei Chen

Electronic Theses & Dissertations (2024 - present)

Detecting and characterizing manipulations in digital media continues to pose a significant challenge within the field of digital forensics. Despite notable advancements, the discipline often remains in a reactive stance against emerging threats. Current state-of-the-art methods, typically evaluated within academic settings, fails to mirror the complexities of real-world disinformation scenarios. These methods generally prioritize high performance based on quantitative metrics, yet they demonstrate a considerable dependency on training data and lack adaptability to new novel attack signatures. With the rapid evolution of attack methodologies, the dependency on highly accurate models that do not generalize or adapt well to unseen threats …

The Response Of Tropopause Phenomena To Stratospheric Variability: Rossby Wave Breaking And Tropopause Polar Vortices, Cameron R. Paquette

Electronic Theses & Dissertations (2024 - present)

Variability in the high-latitude wintertime stratospheric flow, often referred to as the stratospheric polar vortex, can precede anomalous conditions in tropospheric flow on subseasonal-to-seasonal (S2S) timescales and impact surface weather. The range of stratospheric impacts on tropospheric conditions range from shifts in the North Atlantic storm tracks, formation of cold spells, and development of extreme precipitation. Impacts of stratospheric variability are known to persist for up to 60 days, providing ample windows of opportunity in forecasting on S2S timescales. This dissertation focuses on the stratospheric impacts on tropopause-level processes, such as Rossby wave breaking (RWB) events and tropopause polar vortices …

Demonstration And Assessment Of Dark Field And Phase X-Ray Computed Tomography, Wadiah Hamed Allahyani

Electronic Theses & Dissertations (2024 - present)

Conventional computed tomography produces a 2D slice or 3D image from multiple conventional images to avoid overlapping structures and increase visualization of the internal features of the object. A set up, macro controls and processing for computed tomography was created for the Center for X-ray Optics for the first time, and demonstrated in this work. Adding differential phase contrast to conventional x-ray images highlights edges. Two more channels of information, dark field and integrated phase, improve the contrast and increase the signal-to-noise ratios of the image. Combining these channels with computed tomography is an important development. This work is the …

Further Results On Learning Quantum Measurement Classes: Quantum Pac Model For Povm Hypothesis Classes, Arka Prabha Das

Electronic Theses & Dissertations (2024 - present)

This thesis investigates the problem of learning from quantum systems, where each example consists of a quantum state paired with a classical outcome. The task centers on choosing an effective measurement rule from a fixed set to enable accurate prediction of the classical outcome from the quantum state. A central focus lies in understanding whether joint measurement strategies that cannot be separated into local operations offer a real benefit in terms of the number of examples needed for successful learning. We examine conditions under which a non-separable measurement within a given hypothesis class achieves strictly better sample complexity bounds compared …

Studies On Convexity Of Dnf Formulae, Josue A. Ruiz

Electronic Theses & Dissertations (2024 - present)

In this dissertation, we investigate the problem of determining whether a Boolean formula given in disjunctive normal form (DNF) is convex. Although Boolean formulas have various applications, our research focuses on the practical application for rule-based access control policies, where policies are often expressed as a set of Boolean rules. Understanding the structural properties of such formulas is crucial for determining whether a policy can be efficiently represented within a specific access control model.

The main contribution of this research is the conception and analysis of convexity derived from the “gap problem.” In this context, convexity is characterized by the …

Light-Driven Chemistry: Photoswitch In Rna Activity Control And Development Of Photoredox Catalysis, Juncheng Li

Electronic Theses & Dissertations (2024 - present)

This dissertation is structured into two principal segments. The initial segment elucidates a series of synthetic works involving artificial molecules, specifically featuring the photosensitive azobenzene. These molecules facilitate remote control over the structure and functionality of biomolecules, with subsequent applications in regulating biological processes. The latter segment concentrates on the design and enhancement of organic photocatalysts, utilizing traditional acridine ions as a foundation. This section details the development of diverse organic photocatalyst-promoted processes, including photo Lewis acid catalysis, photo Brønsted acid catalysis, and photoredox reactions.

Issues And Challenges In Silicon Based Quantum Computing, Rafia Ayub

Electronic Theses & Dissertations (2024 - present)

Silicon-based quantum computing has emerged as a promising platform for scalable and fault-tolerant quantum information processing. This thesis investigates the use of silicon quantum dots as qubits, addressing a key limitation in their implementation—charge noise and decoherence. We explore the physical and electronic properties of silicon quantum dots, focusing on their coherence times, tunability, and compatibility with existing semiconductor fabrication technologies. Through theoretical analysis and numerical simulations, we examine the impact of material imperfections and propose strategies to mitigate decoherence effects, thereby enhancing qubit stability. Additionally, we discuss potential pathways for integrating silicon quantum dots into large-scale quantum architectures. Our …

Al(X)In(Y)Ga(1−X−Y)N-Based Monolithically Integrated Ir Detector-Visible Emitter Device, Alireza Lanjani

Electronic Theses & Dissertations (2024 - present)

III-Nitride material system has been widely used in electronic, sensing, optoelectronic, and power device applications. Owing to the wide and tunable bandgap of this material system and the realization of relatively high conductivity p-type Mg-doped GaN films, considerable research over the past few decades has focused on applications based on wide bandgaps, such as blue and ultraviolet (UV) light-emitting diodes (LEDs), laser diodes, visible-blind and solar-blind photodetectors. The other side of the spectrum, infrared (IR) has remained considerably less explored.

To exploit III-Nitrides for IR detection applications, the quantum well infrared photodetector (QWIP) design is employed. In the QWIP …

Misoscale Vortices In The Eyewall Boundary Layer Of A Tropical Cyclone In A Large Eddy Simulation, Emily A. Lucy

Electronic Theses & Dissertations (2024 - present)

Since the late 1980s, there have been numerous observations of extreme updrafts, gusts, and small-scale rotation in the eyewall of tropical cyclones (TCs). Research aircraft, dropsondes, and, more recently, mobile radars have collected these observations, most often from within the boundary layer. Previous studies have documented dropsondes associated with extreme gusts and updrafts, while others have discussed specific incidents of misoscale features in tropical cyclones. Several high-resolution numerical simulations of TCs have documented similar features. This study uses one such simulation from Worsnop et al. (2017). A small but intense tropical cyclone in a quiescent environment is simulated.

This simulation …

Dynamics Of Mesoscale Offshore Wind Flows From An Observational And Modeling Perspective: Implications For Offshore Wind Energy, Elizabeth J. Mccabe

Electronic Theses & Dissertations (2024 - present)

Mid-latitude coastal regions of the United States, such as the New York Bight (NYB), are playing a major role in the development of offshore wind energy. However, mesoscale dynamics of local and regional circulations in these offshore waters are poorly understood. The sea breeze, especially in the NYB, will play a key role in the development of offshore wind energy, given the favorable wind speeds and high capacity factors during periods of peak demand. With limited vertical wind profile observations available offshore, the ability to accurately forecast and understand the spatiotemporal extent of the sea breeze and often associated low-level …

Food Rescue, Food Waste, And Policy In New York: A Mixed-Methods Study Of Food System Impacts, Mariana Torres Arroyo

Electronic Theses & Dissertations (2024 - present)

This dissertation investigates how food donation policies in New York State influence the recovery, redistribution, and waste of surplus fresh produce, with a focus on fruits and vegetables. Using an interdisciplinary, systems-based approach that integrates system dynamics modeling with qualitative, community-engaged research, the study analyzes the effectiveness and unintended consequences of three major policies: Nourish New York, the Farm to Food Bank tax credit, and the Food Donation and Food Scraps Recycling Law.

Chapter 2 uses simulation modeling to examine how the tax credit and waste ban affect produce recovery and redistribution. Results show that while both policies can boost …

Betti Numbers Of Generic Ideals, Jason R. Howell

Electronic Theses & Dissertations (2024 - present)

We present results related to Betti numbers of so called generic ideals over a polynomial ring $T = \Bbbk[x_1, \ldots, x_n]$. For each $m$ define $T(m) = T/(x_{m+1}, \ldots, x_n)$ and for a homogeneous ideal $J$ we use the notation $J(m) = JT(m)$. Also we set $Q(m) = T(m)/J(m)$, and $L(m) = ann_{Q(m)}(x_m)$. The first main result is Theorem \ref{Long Exact Sequence} where we produce the following long exact sequence \begin{align*} \cdots \rightarrow &Tor_{k+1,j}^{T(m-1)}(Q(m-1),\Bbbk) \rightarrow Tor_{k-1,j+1}^{T(m-1)}(L(m),\Bbbk)_{j-1} \rightarrow Tor_{k,j}^{T(m)}(Q(m),\Bbbk) \rightarrow \\ &Tor_{k,j}^{T(m-1)}(Q(m-1),\Bbbk) \rightarrow Tor_{k-2,j-1}^{T(m-1)}(L(m),\Bbbk) \rightarrow \cdots. \end{align*} The second main result is Theorem \ref{Theorem F(j,m) equiv k(j,m)} where we explicitly describe …

Comparison Of Methane Emission Quantification Methods At Landfills In New York State, Alexandra M. Catena

Electronic Theses & Dissertations (2024 - present)

Greenhouse gas (GHG) emission increases are a consequence of global population rise and economic growth. It is well known that carbon dioxide is the dominant GHG by mass, but methane has received increasing attention recently due to its high warming potential and relatively short lifetime of about a decade. Methane has several different types of both natural and anthropogenic sources including wetlands, waste management, energy, agriculture and farming, and wildfires, which makes tracking and reducing methane emissions a complex problem. The emission rates of each of these sources are highly uncertain due to differing emission estimation methods, which can be …

Backgrounds And Efficiencies For Low-Energy Searches, And Improved Calibration Techniques In The Lux-Zeplin Dark Matter Experiment, Gregory M. Blockinger

Electronic Theses & Dissertations (2024 - present)

The full understanding of dark matter has eluded physicists for many years. While there is abundant astrophysical evidence across many scales, the exact nature of this matter, which comprises about 25% of the Universe’s energy density, remains unknown. Weakly Interacting Massive Particles (WIMPs) are one of the leading theoretical candidates, and numerous ex- periments have searched for their interactions. In recent decades, dual-phase Time Projection Chambers (TPCs) have become the leading technology in this search. The LUX-ZEPLIN (LZ) detector is one such xenon-based experiment, operating nearly a mile underground in the Black Hills of South Dakota at the Sanford Underground …

Boundary Conditions, Symmetries, And Bootstrap In 2d Conformal Field Theory, Yucong Cai

Electronic Theses & Dissertations (2024 - present)

This dissertation contains two parts of study on the Friedan states and on the conformal bootstrap with symmetries.

The Friedan states are potential physical boundary states of the free boson. However, they exhibited a continuous spectrum in the open string sector, in contrast to more standard examples. The explicit expressions for the density of states of the Friedan states is obtained. Some pathologies and possible contradictions of these states are explored. The $g$ functions of the Friedan states are shown to be infinite, suggesting an infinite number of degrees of freedom in the theory.

The modular conformal bootstrap using the …

Identification Of Fentanyl And Fentanyl Analogs Found In Human Urine Using Gas Chromatography-Mass Spectrometry, Adam M. Fisher

Electronic Theses & Dissertations (2024 - present)

Identification of fentanyl and its analogs in unknown samples is a frequent practice completed in forensic crime labs, as the fentanyl epidemic continues. With the continued increase in illicit drug use, crime labs are not able to keep up with the flooding of new fentanyl analogs being developed, resulting in new analogs slipping past the crime labs’ authority. This research is to start to develop a database of commonly found fentanyl analogs in New York State. Gas Chromatography-Mass Spectrometry (GC-MS) is a widely used method for the detection and identification of illicit drugs in forensic crime labs. A database that …

Improving Generalizability In Image Manipulation Detection, Zhenfei Zhang

Electronic Theses & Dissertations (2024 - present)

Image manipulation detection (IMD) aims to determine whether an image has been tampered with and to identify the manipulated regions. These capabilities have become increasingly important with the rapid advancement of media editing and generation technologies, such as Photoshop and generative AI methods, which underscore the need for robust tools for media authentication. Although current state-of-the-art (SoTA) methods achieve strong results on common manipulation types, such as splicing, copy-move, and removal, they often struggle to generalize to manipulation types not represented in the training data. Consequently, their real-world applicability remains limited, with performance degrading significantly in practical scenarios.

In this …

Knowledge And Ontology Enhanced Approach To Natural Language Understanding (Koe-Nlu) In Computational Social Media And Healthcare, Naga Usha Gayathri Lokala

Theses and Dissertations

Natural Language Understanding (NLU) faces both opportunities and challenges as the amount of social media and healthcare data grows. This is particularly evident in context-sensitive applications such as evaluating cognitive health, identifying mental health symptoms, and monitoring drug abuse. Even though traditional NLU models work well for processing language in a wide range of areas, they often lack the ability to understand language in a specific domain, reason in context, and incorporate structured external knowledge. This dissertation talks about the Knowledge and Ontology Enhanced Approach to Natural Language Understanding (KOE-NLU), a new framework that is meant to make NLU systems …

Stereoselective Synthesis Of (Z)-Acrylic Nitriles From The Ruthenium-Catalyzed Coupling Reaction Of Nitriles With Unsaturated Carbonyl Compounds Via C–C Bond Cleavage, Aldiyar Shakenov, Krishna Prasad Gnyawali, Chae S. Yi

Chemistry Faculty Research and Publications

Acrylic nitriles are a versatile class of synthetic precursors for a variety of pharmaceutically active compounds, as well as for nitrile polymers. We devised a stereoselective synthesis of (Z)-acrylic nitriles from the Ru-catalyzed coupling reaction of nitriles with unsaturated carbonyl compounds via C–C bond cleavage. Both carbon KIE and Hammett correlation data indicated that C–C bond cleavage is the rate-determining step for the coupling reaction. Several bioactive (Z)-acrylic nitriles were synthesized by using the catalytic coupling method.

Human Perception Of Ai Capabilities At Classifying Perturbed Roadway Signs, Katherine R. Garcia, Jing Chen, Yanru Xiao, Scott Mishler, Cong Wang, Bin Hu

Psychology Faculty Publications

Artificial Intelligence (AI) is crucial to numerous functions required for driving automation systems, including the computer vision techniques used to detect the roadway environment and make real-time decisions. However, the images used as inputs to the AI system may be maliciously perturbed, or manipulated, causing the AI system to make an incorrect classification. In this study, we examined humans’ perception of the AI’s computer vision capability of classifying various road sign images, including the original images, images with two different types of malicious attacks, and images that are scrambled randomly at the pixel level. Our results showed that participants rated …

Automation To Autonomy: Temporal Dynamics Of Trust And Visual Attention Allocation Did Not Evolve, Tetsuya Sato, Eric Chancey, Yusuke Yamani

Psychology Faculty Publications

Emerging work environments are expected to implement autonomy that performs various functions without human input. Previous works has shown that trust in automation is negatively correlated with visual attention allocation, indicating that trust is a dynamic construct. Moreover, trust in automation and trust in autonomy appears to evolve in similar ways. However, recent work has demonstrated differences between trust in automation and trust in autonomy within Kaber’s (2018) theoretical framework (Sato et al., 2023b). Yet, it is uncertain whether the development of trust and visual attention allocation differs between automation and autonomy. The present study examined the temporal dynamics of …