Physical Sciences and Mathematics Commons^™

Towards Explainable Neural Network Fairness, Mengdi Zhang

Dissertations and Theses Collection (Open Access)

Neural networks are widely applied in solving many real-world problems. At the same time, they are shown to be vulnerable to attacks, difficult to debug, non-transparent and subject to fairness issues. Discrimination has been observed in various machine learning models, including Large Language Models (LLMs), which calls for systematic fairness evaluation (i.e., testing, verification or even certification) before their deployment in ethic-relevant domains. If a model is found to be discriminating, we must apply systematic measure to improve its fairness. In the literature, multiple categories of fairness improving methods have been discussed, including pre-processing, in-processing and post-processing.
In this dissertation, …

Assessing The Impact Of An Intervention To Build Food Agency During Emerging Adulthood, Amy Finley

Graduate College Dissertations and Theses

Agriculture is an ancient human activity that has always changed the Earth. But the scale and scope of modern, industrialized agriculture is producing emergent problems in the food system, like unprecedented environmental degradation and high-calorie nutrient-poor diets that are driving poor health outcomes. Increasing cooking skills and the frequency of home cooking have been proposed as solutions to escalating rates of nutrition-related public health problems and may also be important to meeting diet-related sustainability goals like reducing meat consumption. Subsequently, the overarching aim of this thesis is to situate the importance of cooking as a strategy for food systems transformation, …

Effective Drag Coefficient Prediction On Single-View 2d Images Of Snowflakes, Cameron Hudson

Graduate College Dissertations and Theses

The drag coefficient of snowflakes is an crucial particle descriptor that can quantify the relationships with the mass, shape, size, and fall speed of snowflake particles. Previous studies has relied on estimating and improving empirical correlations for the drag coefficient of particles, utilizing 3D images from the Multi-Angled Snowflake Camera Database (MASCDB) to estimate snowflake properties such as mass, geometry, shape classification, and rimming degree. However, predictions of the drag coefficient with single-view 2D images of snowflakes has proven to be a challenging problem, primarily due to the lack of data and time-consuming, expensive methods used to estimate snowflake shape …

Robust Interventions In Network Epidemiology, Erik Weis

Graduate College Dissertations and Theses

Which individual should we vaccinate to minimize the spread of a disease? Designing optimal interventions of this kind can be formalized as an optimization problem on networks, in which we have to select a budgeted number of dynamically important nodes to receive treatment that optimizes a dynamical outcome. Describing this optimization problem requires specifying the network, a model of the dynamics, and an objective for the outcome of the dynamics. In real-world contexts, these inputs are vulnerable to misspecification---the network and dynamics must be inferred from data, and the decision-maker must operationalize some (potentially abstract) goal into a mathematical objective …

Organic Fouling Mitigation In Forward Osmosis Technology Through The Use Of Oscilatting Alternating Current Electric Fields, Logan Werner

Graduate College Dissertations and Theses

Forward osmosis (FO) is the term given to osmosis in water filtration applications. FO has many advantages to conventional membrane filtration processes. The lack of external pressure needed to force solvent through the membrane is dramatically decreased in FO, resulting in a lower cost of operation compared to reverse osmosis. Lower external pressures also result in decreased fouling on the membrane surface and improved permeate flux. Fouling is one of the foremost challenges within the membrane filtration industry and is one of the biggest contributors to operating costs. While FO results in less fouling than RO, fouling remains a major …

Employing Genetic Algorithms For Energy-Efficient Data Routing In Internet Of Things Networks, Farzana Akhter

Graduate College Dissertations and Theses

The Internet of Things (IoT) connects a vast number of smart objects for various applications,such as home automation, industrial control, and healthcare. The rapid advancement in wireless technologies and miniature embedded devices has enabled IoT systems to be deployed in various environments. However, the performance of IoT devices is limited because of the imbalance of data traffic on different router nodes. Nodes that experience high data volume will have a higher energy depletion rate and, as a result, will reach the end of their life quicker than other routers that have less data traffic. Genetic Algorithms are a well-known technique …

Computing The Canonical Ring Of Certain Stacks, Jesse Franklin

Graduate College Dissertations and Theses

We compute the canonical ring of some stacks. We first give a detailed account of what this problem means including several proofs of a famous historical example. The main body of work of this thesis expands on our article \cite{Franklin-geometry-Drinfeld-modular-forms} in describing the geometry of Drinfeld modular forms as sections of a specified line bundle on a certain stacky modular curve. As a consequence of that geometry, we provide a program: one can compute the (log) canonical ring of a stacky curve to determine generators and relations for an algebra of Drinfeld modular forms, answering a problem posed by Gekeler …

Ring Learning With Errors, Sarah Days-Merrill

Graduate College Dissertations and Theses

Over the last twenty years, lattice-based cryptosystems have gained interest due to their levelof security against attacks from quantum computers. The main cryptosystems are based on the hardness of Ring Learning with Errors (RLWE). The Learning with Errors (LWE) problems were first introduced in 2005 by Regev [Reg09] and in 2010, [LPR10] developed the Ring Learning with Errors (RLWE) problems as candidates for safe encryption against quantum computers. Let K be a number field with ring of integers OK. For a prime q, the RLWE problems rely on samples of the form (a, b) ∈ OK/qOK × OK/qOK where a …

Entrepreneurial Behavior, Technology Adoption, And Climate Change Resilience: Evidence From Small And Medium-Scale Livestock And Maple Producers In Vermont, Dimagi Pitawala

Graduate College Dissertations and Theses

Agriculture, particularly livestock and maple production, is at the core of Vermont’s identity, landscape, and economy. Although Vermont is famous for its unique small farms and exceptional maple syrup, more recently it has become known as one of the fastest-warming states in the United States, threatening the viability of the state’s vibrant agricultural sector. In addition to climate change, increasing farm consolidation, low farmgate prices, market competition, and changing consumer demands and lifestyles have begun to affect the viability of small and medium-scale livestock farms and maple operations. While numerous policies and programs seek to ensure the viability of these …

Probabilistic Modeling Of Disease: Addressing Uncertainties In Within-Host And Population-Level Dynamics, Mariah Boudreau

Graduate College Dissertations and Theses

Mathematical modeling of disease dynamics provides powerful tools to understand, predict, and evaluate emerging diseases. These insights aid public health officials, along with other modelers. With a plethora of models to choose from, it is important to consider a model that encapsulates the stochastic nature of disease dynamics. Stochasticity not only conveys chances of stochastic extinction, but provides probabilistic outcomes, essential for capturing the stochastic nature of the real world. In this thesis, three stochastic models are presented, each addressing uncertainties in mechanisms and interpretation of these models, to aid other modelers and decision makers.Starting with the source of infection …

Stable Oxygen Isotope Variability As A Proxy For Large-Scale Dynamics Of The South American Summer Monsoon, Rebecca Orrison

Electronic Theses & Dissertations (2024 - present)

Terrestrial archives of oxygen isotope records are particularly important for the study of South American hydrology, where hydroclimate observations over the historical period are sparse and often not long enough to characterize multidecadal to centennial-scale changes. Knowledge of such low-frequency changes is critically important for interpreting and constraining projections of future change and detecting the emergence of forced change beyond the envelope of natural variability. This dissertation focuses on the past, present, and future of South American tropical and subtropical climate, with an emphasis on the mature phase of the monsoon season (December, January, and February; DJF) using oxygen isotopes …

Using Airborne Measurements To Evaluate Forecasts Of Freezing Drizzle Aloft Results From The Wintre-Mix Field Campaign, Megan Schiede

Electronic Theses & Dissertations (2024 - present)

Icing associated with the accretion of supercooled large droplets (SLD) serves as a significant hazard to aviation activities. Forecasting SLD events is complex due in part to the limitations of numerical weather prediction models. The Winter Precipitation Type Research Multiscale Experiment (WINTRE-MIX) aimed to comprehend such limitations within numerical forecasts of precipitation type (p-type) through the use of extensive observations during events with multiple p-types. To investigate precipitation processes aloft in winter storms with near-freezing surface conditions, research flights were orchestrated with the National Research Council of Canada Convair-580 aircraft. This study examines a WINTRE-MIX intensive observing period that took …

Agricultural Groundcover Update December 2023, Justin Laycock

Natural resources published reports

Summary

• About 96% of the grainbelt had adequate vegetative groundcover (more than 50%) to prevent wind erosion in December 2023.
• In the northern half of the grainbelt, a larger-than-average area has 51–60% groundcover, which is expected to decrease to below 50% over the summer.
• Just under 4% of the grainbelt (553,000 ha) had less than 50% groundcover, which is inadequate to prevent wind erosion. West Midlands Ag Soil Zone had the highest risk of wind erosion and 11.4% of this farmland had inadequate groundcover.
• Less than 0.5% of the grainbelt had a high to very high risk of wind erosion …

A Corona Theorem For Multipliers On The Dirichlet Space, Alea Wittig

Electronic Theses & Dissertations (2024 - present)

An analogue to Wolff's ideal problem for the multiplier algebra of the Dirichlet space, the main theorem provides sufficient conditions to classify membership of an arbitrary function in an infinitely generated ideal of the multiplier algebra.

Leveraging Redundancy As A Link Between Spreading Dynamics On And Of Networks, Felipe Xavier Costa

Electronic Theses & Dissertations (2024 - present)

A constant quest in network science has been in the development of methods to identify the most relevant components in a dynamical system solely via the interaction structure amongst its subsystems. This information allows the development of control and intervention strategies in biochemical signaling and epidemic spreading. We highlight the relevant components in heterogeneous dynamical system by their patterns of redundancy, which can connect how dynamics affect network topology and which pathways are necessary to spreading phenomena on networks. In order to measure the redundancies in a large class of empirical systems, we develop the backbone of directed networks methodology, …

Advancements In Forensic Analysis: Development Of Mass Spectrometric And Chemometric Approaches For The Identification Of Synthetic New Psychoactive Substances And Plant Materials, Mónica Ventura

Electronic Theses & Dissertations (2024 - present)

While forensic science is a well-established discipline, a number of federal agencies have highlighted challenges that continue to plague the field and the extent to which these challenges remain unaddressed. Examples are the illegal trade of wildlife timber and the drug epidemic. Characterization of these materials requires nuanced method development for compound determination, matrix material-specific protocols, and heavy use of expensive consumables. The application of a technique such as direct analysis in real time – high-resolution mass spectrometry (DART-HRMS) provides the opportunity to circumvent many of the challenges presented by conventional methods. In general, little to no sample preparation is …

A Comparison Of Arctic And Atlantic Basin Cyclone Predictability From A Climatology And Case Study Perspective, Peyton Capute

Electronic Theses & Dissertations (2024 - present)

Arctic cyclones (ACs) can transport warm, moist air into the Arctic region, which combined with strong winds may lead to rapid declines in sea ice during the summer. As a consequence, accurate sea ice predictions during the summer may rely on being able to predict cyclone-related wind speed and direction, which critically depends on the cyclone track and intensity. Despite this, there are relatively few studies that have documented the predictability of ACs during the summer, beyond a few case studies. In addition, there has been no extensive comparison of whether these cyclones are more or less predictable relative to …

Multi-Activity Student Knowledge And Behavior Modeling Via Transfer Learning, Siqian Zhao

Electronic Theses & Dissertations (2024 - present)

Online education systems have grown in popularity over the past few years, providing abundant opportunities for students to learn. As the number of students using these systems grows, it promotes the development of the Educational Data Mining (EDM) field, which leverages statistical, machine learning, and data mining technologies to explore large-scale educational data and develop methods to better understand student learning.

In this dissertation, we investigate two essential topics in EDM: Student Knowledge Tracing (KT) and Behavior Modeling (BM). KT aims to quantify and model student knowledge gained from learning activities, while BM focuses on tasks such as modeling student …

Sparse Representation Learning For Temporal Networks, Maxwell Mcneil

Electronic Theses & Dissertations (2024 - present)

Temporal networks arise in many domains including activity of social network users, sensor network readings over time, and time course gene expression within the interaction network of a model organism. Data of this type contains a wealth of prior information such as the connectivity among nodes (e.g., a friendship graph), and prior knowledge of expected temporal patterns (e.g., periodicity). Modeling these temporal and network patterns jointly is essential for state-of-the-art performance in temporal network data analysis and mining. Sparse dictionary encoding is one modeling approach for such underlying patterns. However, most classical approaches consider only one dimension of the data …

Understanding Local And Large-Scale Drivers Of Rainfall Variability And Change Over The Hawaiian Islands Region, Katrina Fandrich

Electronic Theses & Dissertations (2024 - present)

Hawaiian rainfall variability is strongly influenced by anthropogenic climate change and natural variability (e.g., Pacific Decadal Oscillation (PDO), Pacific North American (PNA) pattern, El Niño–Southern Oscillation (ENSO)). For local decision-makers and resource managers, understanding the drivers of rainfall variability and change in Hawai‘i is crucial for climate adaptation and mitigation planning. Previous studies have applied statistical and dynamical downscaling to produce high-resolution rainfall projections for the Hawaiian Islands region. However, to date, downscaling has only been used to investigate the influence of anthropogenic forcing on late-century rainfall projections for the Hawaiian Islands. For regions like Hawai’i, that are highly vulnerable …

Predictability Of Downslope Windstorms And Mechanisms Controlling Their Onset, Matthew James Brewer

Electronic Theses & Dissertations (2024 - present)

Downslope windstorms in the western United States often contribute to the ignition and ex- acerbation of existing wildfires due to their hot, dry, and gusty nature. Accurate forecasts of these windstorms are imperative for mitigating wildfire risk. However, forecasting downslope windstorms remains challenging due to their sensitivity to environmental conditions.

The Sundowner Windstorms of the Santa Barbara, California region have been asso- ciated with many large wildfires in the area. This study analyzes special observations from the Sundowner Wind Experiment field campaign to better understand the diurnal cycle of these windstorms. Analysis of the strongest windstorm during the campaign indicates …

An Image Processing Approach For Real-Time Safety Assessment Of Autonomous Drone Delivery, Assem A. Abdelhak, Dan Moss, Alan Hicks, Susan Mckeever

Articles

The aim of producing self-driving drones has driven many researchers to automate various drone driving functions, such as take-off, navigation, and landing. However, despite the emergence of delivery as one of the most important uses of autonomous drones, there is still no automatic way to verify the safety of the delivery stage. One of the primary steps in the delivery operation is to ensure that the dropping zone is a safe area on arrival and during the dropping process. This paper proposes an image-processing-based classification approach for the delivery drone dropping process at a predefined destination. It employs live streaming …

The Impact Of The Use Of Mercury In Mining In Colombia, Johan Martinez

CMC Senior Theses

Artisanal and small-scale gold mining is a practice that is common all around the world, especially in Latin America, Africa, and Asia. It is also one of the leading consumers of mercury and leading sources of emissions. Governments have banned the use of mercury completely, including Colombia, but it continues to be used and emitted into the environment. This issue is exacerbated because there is a lack of standards, a lack of research done regarding mercury use in mining, and a complex socio-political history in Colombia. For this reason, this paper seeks to provide a proposed pathway for research to …

Determining The 2022-2023 Mass Balance Of The Palisade Glacier In The Sierra Nevada Mountains Of California With Remote Sensing, Density Modeling, And Temperature-Index Techniques, Vijay Jain, Eric Grosfils, Sarah Kavassalis

CMC Senior Theses

Small, alpine glaciers, such as those in the Sierra Nevada, are difficult to study because of their small size and remoteness, however, they are important recorders of the impacts of climate change in temperate, alpine environments. Previous studies have attempted to characterize the health of these glaciers using extent change techniques, but these methodologies can only roughly approximate the rudimentary measurement of changing ice volume. This thesis uses the Airborne Snow Observatory Inc.’s (ASO’s) aerial lidar snow depth datasets to perform a mass balance calculation for the Palisade Glacier over the remarkable 2022-2023 water year (October 1 through September 30), …

C. Compactum Acts As A Comprehensive Climate Archive And Ecological Foundation In The Labrador Sea, Sadie Heckman

CMC Senior Theses

Clathromorphum compactum, a species of crustose coralline algae (CCA), is incredibly valuable for the future of high latitude ocean health, both as a comprehensive archive of changing ocean conditions, and ecologically as a foundational species for promoting biodiversity. Previous work establishes C. compactum as an effective climate proxy, and its life history provides several advantages for this use. C. compactum grow in nongeniculate, generally radial formations on hard substrates, over a wide distribution in mid-to-high latitude oceans and at subtidal depth ranges. Indeterminate growth leads to extreme longevity in C. compactum (Halfar et al., 2008), and growth rates are relatively …

Soil And Fertiliser Management For Orchards In The Manjimup District, Neil Lantzke, Simon Yeap, Lukasz Kotula, B Goldspink, Valeria Almeida Lima, John Kammann

Horticulture research reports

This publication aims to give orchardists a broad understanding of the soils in the Manjimup district. It provides information on soil and fertiliser management that will assist in the sustainable production of fruit crops. This document was prepared for a Soils Course for horticulturalists, that was held in September 2023 at Manjimup Horticultural Research Centre.

The Manjimup local government area (LGA) is the most important horticultural district in Western Australia, particularly for fruit and in 2020 the total value of horticultural production was valued at $236 million (Figure 1). Avocados and apples were the most valuable horticultural crops, valued at …

Echolocation On Manifolds, Kerong Wang

Honors Theses

We consider the question asked by Wyman and Xi [WX23]: ``Can you hear your location on a manifold?” In other words, can you locate a unique point x on a manifold, up to symmetry, if you know the Laplacian eigenvalues and eigenfunctions of the manifold? In [WX23], Wyman and Xi showed that echolocation holds on one- and two-dimensional rectangles with Dirichlet boundary conditions using the pointwise Weyl counting function. They also showed echolocation holds on ellipsoids using Gaussian curvature.

In this thesis, we provide full details for Wyman and Xi's proof for one- and two-dimensional rectangles and we show that …

Self-Quenching Of Carbon Monoxide In The Presence Of Noble Gases, Madeline V. Hinkle

Honors Theses

Precise rate coefficients for the vibrational relaxation pathways of CO(v)-CO in the presence of Ar and Kr have been determined through the work of this thesis. This work was motivated by the need to find a more cost-effective alternative to using xenon as a bath gas, which has increased significantly in price in the past few years. Similar experiments within this lab at Bucknell have been conducted in the past using argon, which can be used in the same manner as xenon but comes at a much lower price, but the experiments yielded inferior results compared to those using xenon. …

The Precedence-Constrained Quadratic Knapsack Problem, Changkun Guan

Honors Theses

This thesis investigates the previously unstudied Precedence-Constrained Quadratic Knapsack Problem (PC-QKP), an NP-hard nonlinear combinatorial optimization problem. The PC-QKP is a variation of the traditional Knapsack Problem (KP) that introduces several additional complexities. By developing custom exact and approximate solution methods, and testing these on a wide range of carefully structured PC-QKP problem instances, we seek to identify and understand patterns that make some cases easier or harder to solve than others. The findings aim to help develop better strategies for solving this and similar problems in the future.

Centers Of N-Degree Poncelet Circles, Georgia Corbett

Honors Theses

Given a circle inscribed in a polygon inscribed in the unit circle, if one connects all the vertices with line segments we get a family of circles called a package of Poncelet circles, due to its connection to a theorem of Poncelet. We are interested in where the centers of the Poncelet circles can be. Specifically, we have shown that if one of the circles in the Poncelet package is centered at 0, then all of the circles must be centered at 0 as well. This was proven by Spitkovsky and Wegert in 2021 using elliptic integrals but we …