As modern-day economists continue to raise concerns about an increasingly likely economic recession, policymakers should examine recent economic crises to determine potential takeaways or implications for future events. The 2008 financial crisis was a defining global political and economic event. Two significant markets impacted were the United States and the European Union, who pursued divergent financial and monetary policy responses to the crisis despite their close relationship. Most existing scholarship focuses on the institutional differences in policy implementation, analyzing their impact on the overall economic trajectory. However, this thesis found that despite differences in economic indicators, anger leading to instability remained in both the U.S. and the EU, regardless of the current overall economic state. In both regions, anger and upheaval were centered around the unequal economic distribution and inequality levels, which the everyday citizens viewed as only exacerbated by the economic crisis and the resultant policy decisions.

Contrary to current economic analysis, which focuses primarily on overarching macroeconomic factors and statistics to draw conclusions, researchers and policymakers should treat income inequality as a significant outcome and indicator of a crisis response rather than an afterthought. The 2008 financial crisis in both markets showed that despite varying levels of recovery in specific economic indicators, the lack of importance placed on equitable economic distribution led to worldwide political upheaval and created a backlash that still reverberates today.

Economic crisis moments, both in the past years and in the future, should be treated as an opportunity to make progress in reducing income and wealth inequality rather than policymakers simply striving to maintain the status quo. The economic downturn caused by the COVID pandemic in 2020 and its resultant response showed that policymakers still are not adequately emphasizing the importance of equitable policies and are underestimating the importance of long-term political stability in favor of economic metrics.

Discretization of the Schwinger model is a common testbed to calculate observables in low-dimensional QED. Extending this model to be spatially 2D is more unexplored and might uncover interesting behaviors. A lattice model is constructed first for the Schwinger model as done in this paper by Dempsey et. al [1] and then the model is extended into 2D.

Do presidential campaign events influence how people vote? Do campaign events impact polling margins? If not, why would presidential candidates and their campaign teams hold hundreds of campaign events in the fall leading up to Election Day? Political science research about the effect of presidential campaign events on election phenomena is mixed. This study examines the relationship between campaign events and average polling margins, at the state level, and vote margins, at the county level, in the 2024 presidential election. This study finds: (1) support for Hypothesis 1 that in a close race, as reflected in the polls, candidates work harder via campaign events in competitive states with a relatively high number of electoral votes; (2) that with respect to Hypothesis 2, the results did not find clear support that within battleground states, candidates hold campaign events where average polling margins are close. Instead, the study finds that for a few states, specifically Michigan, Wisconsin, and perhaps Pennsylvania, there is a moderate to weak correlation that there are fewer campaign events when the margin is larger. However, when the individual battleground state regressions were performed, generally in most states, there was no statistically significant relationship between polling margins and the number of campaign events in the week before the poll or the week following the poll, with few exceptions. With respect to Hypothesis 3, the study did not find clear support that the number of campaign events in a given county had a positive affect on vote margin at the county level. Instead, the regression results may have been affected by selection bias or the tight one-week time frame applied to measure the data.

Soft mobile robots offer distinct advantages for navigating complex terrains because of their inherent flexibility, which enables exceptional adaptability and versatility. However, their compliant bodies introduce significant challenges, such as motion uncertainties and unpredictable interactions with their environment, that are difficult to control. Furthermore, the complex dynamics of soft mobile robots complicate the realisation of full autonomy, a challenge that is further exacerbated by the limited sensing and proprioception capabilities employed in the field.

This thesis aims to advance the field of autonomous origami-enabled mobile robots, a subclass of soft mobile robots, by enhancing their capability to traverse complex terrains and improving their viability for real-world applications. Previous work from the Bio-inspired Adaptive Morphology Laboratory (BAM Lab) developed an Origami-Enabled Soft Crawling Autonomous Robot (OSCAR) in pursuit of this goal. OSCAR is a novel soft mobile robot that leverages origami-inspired mechanisms to mimic the crawling motion of caterpillars. Hence, building upon that foundation, this work enhances OSCAR’s capabilities by enabling traversal of complex three-dimensional spaces without relying on external sensors, paving the way for implementation of truly autonomous navigation.

OSCAR’s mechanical stability is first enhanced with a double-celled design, and vertical actuation is introduced by employing four origami towers per cell. These upgrades improve stability, maneuverability, and locomotion range, enabling complex three-dimensional terrain traversal in the updated version called the Slinky Origami-Enabled Soft Crawling Autonomous Robot (SOSCAR). Afterwards, control systems are developed to realise the new mechanical design and demonstrate vertical obstacle avoidance. Finally, internal sensing mechanisms, using Time-of-Flight distance sensors and Inertial Measurement Units, are integrated to provide proprioception, or self-awareness, that enable closed-loop positional feedback control. Whereas previous versions of OSCAR relied on external sensors for control, all sensing in SOSCAR is fully integrated onboard the robot.

Ultimately, this thesis presents a soft mobile robot that integrates the necessary elements for future implementation of autonomous navigation in complex three-dimensional terrains. As a result, it advances the real-world readiness of origami-enabled robots and highlights their potential for operating in challenging environments.

This research paper investigates the impact of “amenity migration” on the local populations in rural mountain towns during the COVID-19 pandemic as well as potential barriers to healthcare that arise when the infrastructure of these communities is strained by a large and sudden population boom. It aims to explore the distinct perspectives, challenges, and adaptations of local residents and healthcare providers by employing interviews as the primary method. The study delves into motivations to live in the mountains, housing limitations, remote work, community dynamics, and balance between full-time and part-time residents during the pandemic. Furthermore, this study takes an analytical approach to theories of social ambivalence as a means to understand the complex and contradictory emotions local populations may experience towards amenity migrants. The social determinants of health play a significant role in analyzing the needs of these communities as they continue to undergo change because of these in-migrants. This paper contributes to the existing literature by filling gaps in understanding the nuanced responses of locals, healthcare workers, ski patrollers, and mountain rescuers to the pandemic in rural settings, offering insights into their distinct coping strategies, perceptions, and contributions to community well-being. This paper aims to identify the existence of “amenity migrants'' in rural mountain towns proposed by previous literature, understand their impact on local populations during the COVID-19 pandemic, and how the health landscape has been altered in the equitable distribution of healthcare. In bringing attention to this phenomenon, I aim to further inform policies and interventions that will support the diverse needs of both the locals and amenity migrants in the areas faced with the challenges posed by the pandemic.

The use of synthetic datasets for computer vision is a key factor in the improvement of many methods. Infinigen is a procedural generator of synthetic 3D scenes of the natural world with the goal of creating datasets for computer vision research. One such area of research is Visual SLAM models, which seek to map an unknown environment while simultaneously tracking an agent’s pose. I propose a Visual SLAM benchmark based on Infinigen as well as an open-source method to generate more data for SLAM algorithms using Infinigen. The Infinigen SLAM benchmark contains extremely challenging camera motion within various indoor environments. State-of-the-art Visual SLAM models perform well on the proposed benchmark, however they perform worse than on comparable SLAM benchmarks. This suggests that Infinigen is capable of producing useful data for future SLAM research.

This thesis presents a user friendly approach to information driven antibody-antigen folding and docking.

The following study investigated temporary declines in air pollution observed during the COVID-19 lockdown and potential changes in adult asthma prevalence across four urban U.S. East Coast cities: Boston, Massachusetts; Baltimore, Maryland; New York City, New York; and Washington, D.C. Leveraging the COVID-19 lockdown as a natural experiment, this research combined annual air pollutant data from the U.S. Environmental Protection Agency (EPA) with asthma prevalence data from the Center of Disease Control’s Behavioral Risk Factor Surveillance System from the years 2013 to 2023 and conducted multiple statistical analyses to determine associations between air pollutants and asthma prevalence. Regression and mediation models were utilized to assess the effects of nitrogen dioxide (NO2) and fine particulate matter (PM2.5) on asthma prevalence. Results showed that NO2 concentrations significantly declined during the COVID-19 period and were modestly associated with reduced asthma prevalence. Mediation analysis suggested that NO2 partially mediated the relationship between the COVID-19 period and observed decreases in asthma prevalence. PM2.5 did not demonstrate a significant mediating effect. Although asthma prevalence declined during the COVID-19 lockdown period, the decrease was not statistically significant. Additionally, the rebound in asthma prevalence after COVID-19 may reflect diagnostic delays or other non-environmental factors impacting the data. Nonetheless, these findings reinforce and inform previous research regarding the impact of air pollution on respiratory health outcomes and asthma prevalence. Furthermore, the results of this study support tactics to reduce NO2 emissions as a strategy to improve public health in urban environments.

Cuticular hydrocarbons (CHCs) serve important roles in insects as barriers against desiccation and as chemical signals within and between species. In eusocial species such as Bombus impatiens, CHCs are known to be correlated with reproductive status, behavior, and caste, but the composition of chemical profiles across different body parts and between castes remains poorly understood. This study investigates caste-specific differences in the cuticular hydrocarbon profiles of the common eastern bumble bee, Bombus impatiens, focusing on the compound classes of alkanes, alkenes, esters, ketones, aldehydes, and terpenoids. I hypothesized that queens would exhibit higher overall levels of CHCs, particularly alkanes, alkenes, and esters, and that the thorax and abdomen would show the most pronounced differences due to their association with the biosynthesis of hydrocarbons. Small colonies were established and aged under controlled conditions, after which queens and workers were dissected and separated by antennae, head, thorax, abdomen, and legs. Pentane extractions of each body part were analyzed via gas chromatography-mass spectrometry. Statistical comparisons using NMDS and ANOSIM revealed strong caste-specific differentiation across all body regions, with queens exhibiting significantly greater abundances of alkanes across all tissues and enriched levels of alkenes and esters on most body parts. The thorax displayed the greatest chemical divergence between queens and workers. The results of this experiment reinforce the idea that reproductive roles strongly shape the composition of cuticular hydrocarbons in bumble bees. Understanding these differences in chemical profiles enhances broader comprehension of eusocial organization and communication and provides further insight into potential mechanisms by which queens maintain reproductive dominance.

This study uses natural language processing (NLP) techniques to analyze the syntactic and lexical differences between Standard French and Cameroonian French, as well as examine how the dialect evolves when used by the Cameroonian diaspora in France. The central methodology involves training and evaluating two distinct NLP models: one fine-tuned on a corpus of Standard French, and the other on Cameroonian French. The LSTM model, on the other hand, outperformed the Logistic Regression model in all key metrics, including accuracy, precision, recall, and F1-score. The results of this study illustrate the limitations of traditional NLP methods, such as logistic regression, when applied to dialects with syntactical and linguistic differences, and they highlight the potential of deep learning approaches to better handle these variations. The findings point to the importance of fostering linguistic diversity within computational models.

Past studies of turbulence in the Interstellar Medium (ISM) have simulated turbulent sources through the Fourier space driving (FSD). These studies have demonstrated that the resulting statistical distributions of the fluid are highly sensitive to the exact driving form, highlighting a need to analyze which form reproduces which aspects of the ISM. Real observations have shown regions of expanding bubbles commonly sourced by supernovae dominate the dynamic structure. This local source of turbulence is largely different from the global scale FSD method. To find regions in the FSD method that best reproduce the statistical distributions created by expanding bubbles, we compare both distributions outputted from MHD simulations. We utilized AthenaK, to which we’ve added a momentum bubble injection method. Our results show no FSD model is able to reproduce the resulting distributions of the momentum injection method. The velocity distributions are largely different between the two methods, with the momentum injection method generating larger power in the velocity field than all tested FSD models. We did find that both the purely compressive and momentum injection methods produce density distributions that are not log-normal. The momentum injection method possibly fits the log-normal distribution well largely, only deviating in low density non- Gaussian portions.

The residential building sector is a major contributor to energy consumption and greenhouse gas emissions, making electrification and intelligent energy management essential for decarbonization. However, increased electricity demand can strain the power grid, leading to higher costs and emissions. Demand-side flexibility, enabled by on-site power generation, energy storage, and optimized control algorithms, can mitigate this problem by shifting electricity consumption to times when electricity is cheaper and cleaner.

This study evaluates three methods for centralized building energy storage management using CityLearn, an open-source environment for simulating and benchmarking building energy control. The evaluation compares Model Predictive Control (MPC) with two Reinforcement Learning (RL) methods: Soft Actor-Critic (SAC) and Proximal Policy Optimization (PPO). The methods are assessed across three dimensions: (1) energy performance, including cost, carbon emissions, electricity consumption, and stability of electricity use over time; (2) computational efficiency, including training time, memory usage, and inference speed; and (3) scalability, measured across different district sizes of two, four, and eight buildings.

Overall, SAC achieved the strongest performance on cost and energy metrics, performing slightly better than PPO in those areas. PPO, however, produced smoother control behavior with more stable electricity use over time while requiring significantly less memory than SAC and less computation than MPC. Both RL methods outperformed MPC across most metrics, with MPC particularly struggling to scale. Nonetheless, MPC remained more interpretable and required no training data, though it involved substantial engineering effort to develop an accurate system model.

These findings highlight trade-offs between performance, stability, and deployability. PPO emerged as the most balanced controller, offering strong performance with scalability and computational efficiency, making it well-suited for real-world use.

Verifying the dynamical similarity between data-drained deep learning models and the biological circuits they aim to replicate remains a significant challenge. Benchmarking methods that evaluate such models based on their underlying dynamical systems, rather than only their output performance, are thus highly desirable. In this work, we validate and extend one such recently proposed method: the Computation through Dynamics Benchmark (CtD-B). In the context of models that solve the Poisson-clicks task (a perceptual decision-making cognitive task), we test existing CtD-B metrics and find that functional similarity measures — Rate R2 and Dynamical Systems Alignment (DSA/co-BPS) — are robust across models, but representational metrics — State R2 and Cycle-Convergence (Cycle-Con) — are reliable for low-dimensional models. Leveraging dynamical systems theory, we extend the analysis function of the benchmark to consider local similarity: fixed points and timescales in both task-trained (TT) and data-driven (DD) models. Notably, we find that DD models can fit observed data without preserving the characteristic timescales of TT solutions.

The International Moth is a small racing sailboat that can reach top speeds of 35 knots (18 m/s), due to its use of hydrofoils, which lift the entire hull clear of the free surface. The hydrofoils replace the hull as the primary generators of hydrodynamic forces within the vessel system, and in turn, heavily drive the overall performance of the vessel. Optimizing the shape and planform of the foils is a key to achieving race-winning designs. However, hydrofoiling sailboats are highly coupled systems that operate in two simultaneous fluid media, and a change in foil configuration can have cascading effects on the overall vessel state. Thus, a design framework is formulated that allows foil designs to be evaluated within a 6 degree of freedom velocity prediction program (VPP). The framework integrates gradient-based shape optimization tools in 2 and 3 dimensions. Evaluation of the framework demonstrates functionality for design optimization independent of the VPP, but the presented approach to modeling hydrodynamic forces within the VPP requires improvement in order to produce meaningful results that can inform design decisions.

Sleep has a profound impact on numerous cognitive functions, including decision-making and self-control. However, the precise mechanisms by which sleep influences these processes are not completely understood. This study explored how sleep loss impacts the weighting of short-term and long-term rewards, and thereby influences the decision-making process. Participants self-reported their sleep on the previous night and participated in a decision-making experiment involving 25 choices between pairs of food items, where the taste value of each food item represented short-term reward and the health value of the food represented long-term reward. I hypothesized that participants who slept less would exhibit less self-control and thus would place a lesser weight on health relative to taste; further, I hypothesized that self-control would deplete as trials progress.

I developed four nested computational models to characterize the decision-making process: a baseline model that assumes equal weights for taste and health, a model that fits a health weighting parameter for each participant, and two models that incorporate a linear or exponential decay parameter to simulate potential self-control depletion across trials. The model that fit a βhealth weight for each participant provided the best fit for the data, suggesting that participants vary in how they weigh taste versus health and that this weighting was relatively stable across trials. This study did not find a significant correlation between βhealth and sleep hours under any of the models. While the results did not align with my hypotheses, this may be due to a small sample size, limited variability in sleep duration, computational constraints, and other factors. Further research is needed to better explore this relationship, potentially with more extreme manipulations of sleep or the consideration of additional factors that may influence intertemporal choice.