Masters Theses

Permanent URI for this collection

Browse

Recent Submissions

Now showing 1 - 20 of 22776
  • A Machine Learning-Based Heuristic to Explain Game-Theoretic Models
    Baswapuram, Avinashh Kumar (Virginia Tech, 2024-07-17)
    This paper introduces a novel methodology that integrates Machine Learning (ML), Operations Research (OR), and Game Theory (GT) to develop an interpretable heuristic for principal-agent models (PAM). We extract solution patterns from ensemble tree models trained on solved instances of a PAM. Using these patterns, we develop a hierarchical tree-based approach that forms an interpretable ML-based heuristic to solve the PAM. This method ensures the interpretability, feasibility, and generalizability of ML predictions for game-theoretic models. The predicted solutions from this ensemble model-based heuristic are consistently high quality and feasible, significantly reducing computational time compared to traditional optimization methods to solve PAM. Specifically, the computational results demonstrate the generalizability of the ensemble heuristic in varying problem sizes, achieving high prediction accuracy with optimality gaps between 1--2% and significant improvements in solution times. Our ensemble model-based heuristic, on average, requires only 4.5 out of the 9 input features to explain its predictions effectively for a particular application. Therefore, our ensemble heuristic enhances the interpretability of game-theoretic optimization solutions, simplifying explanations and making them accessible to those without expertise in ML or OR. Our methodology adds to the approaches for interpreting ML predictions while also improving numerical tractability of PAMs. Consequently, enhancing policy design and operational decisions, and advancing real-time decision support where understanding and justifying decisions is crucial.
  • The Effects of Mechanical Site Preparation Treatment and Species Selection on Survival and Carbon Pools in 12-Year-Old American Sycamore (Platanus occidentalis) and Willow Oak (Quercus phellos) Riparian Plantations in the Southeastern U.S. Piedmont
    Lynn, Drake Havelock (Virginia Tech, 2024-07-16)
    Riparian wetlands may provide numerous ecosystem services, including water quality protection, food and fiber supply, wildlife habitat, and carbon sequestration. In recent years, riparian forests have received increased attention and funding for riparian forest restoration projects. Unfortunately, failures of riparian restoration efforts are likely due to mortality of planted trees. Tree mortality is commonly attributable to several factors, including selection of species that are not well suited to the wetland sites, inadequate planting densities, soil compaction associated with former agricultural activities, lack of microtopographic relief that allow small seedling to survive on wet sites, competition by herbaceous plants, and browse. Selection of well-suited species, dense planting and use of mechanical soil site preparations are all potential remedies to partially address success of wetland restoration plantings. Riparian restoration projects have historically been undertaken with goals of improving water quality and/or wildlife habitat, but in recent years there has been increased valuation of carbon sequestration. Carbon valuation appears to be increasing, but more research is needed to determine rates and pools of carbon accumulation in riparian areas. Our research quantifies forest establishment effects on multiple carbon pools in a densely planted, 12-year-old old-field riparian restoration. Our research evaluated the effects of four soil mechanical site preparations (bed, disk, pit, and mound and rip) and species selection (American sycamore (Platanus occidentalis) and willow oak, (Quercus phellos) on forest establishment and carbon storage across multiple pools, namely in planted trees, herbaceous vegetation, fine roots, organic soil horizons, and the mineral soil. At 12 years, we found that species selection was more important to carbon storage than site preparation. American sycamore was well suited to the site and had better survival than willow oak (64% vs 42% survival). American sycamore also stored more carbon across all site preparations than willow oak. Measured carbon storage averaged 74.8 Mg ha-1 for American sycamore treatments and 63.1 Mg ha-1 for willow oak treatments. The plots were densely planted (1.2 m (4ft) by 1.8 m (6ft) spacing), and forests were established even in higher mortality willow oak plots. These results indicate that high planting density is potentially a viable practice for establishing riparian forest cover, especially if desired species are marginally site suited or other survival inhibiting factors exist.
  • Elucidating the Function of Krüppel Homolog 1 (Kr-h1) Associated Proteins (KAPs) in Aedes aegypti Reproduction Through RNA Interference-Mediated Downregulation
    Zhang, Liyan (Virginia Tech, 2024-07-15)
    The transcription factor Krüppel homolog 1 (Kr-h1) is crucial in multiple reproductive processes of Aedes aegypti mosquitoes, including previtellogenesis, vitellogenesis, and oogenesis. This study explores the interaction between Kr-h1 and its potential associated proteins (KAPs), with a specific focus on the dimerization partner (DP-1), and how this interaction regulates gene expression pathways critical for mosquito reproduction. Utilizing RNA interference (RNAi), the research identifies DP-1 as a significant regulator of follicle growth post-eclosion (PE), highlighting its vital role in the mosquito reproductive regulatory pathway. The experimental approach included RNAi-mediated knockdown of DP-1, accompanied by evaluations using quantitative PCR (qPCR), Western blotting (WB), co-immunoprecipitation (Co-IP), follicle length measurement, and egg counting to assess the role of DP-1 in reproductive functions. For the first time, the inhibition of DP-1 expression was found to significantly impede A. aegypti follicular development. The elucidation of the mechanistic roles of Kr-h1 and DP-1 provides valuable insights that could lead to innovative strategies for mosquito population control and effective disease vector management.
  • Influence of Terrain, Vegetative, and Hydraulic Properties on Sediment and Microplastic Accumulation in the Stroubles Creek Floodplain
    Smith, Tyler Camden (Virginia Tech, 2024-07-11)
    Sediment and microplastic accumulation in stream systems occur when particulates entrained in overbank flow are retained by the floodplain. Despite overbank flow conditions dictating sediment and microplastic accumulation, the spatial distribution of accumulation within floodplains remains poorly understood. Difficulty quantifying hydraulic flow conditions is due to spatial variation introducing erroneous error. This study hypothesized floodplain sediment and microplastic accumulation would be closely correlated with topographic, vegetative, and hydraulic conditions. To test this hypothesis, sediment and microplastic accumulation were measured along a 1.25 km stretch of Stroubles Creek in Blacksburg, Virginia. Sediment accumulation was measured using tiles with a surface area of 144cm² at 75 locations. Tiles accumulated 4,782g over their two-year deployment. Microplastic accumulation was assessed by taking 15cm3 soil grab samples from 40 locations. Microplastics were identified using FTIR spectroscopy and were found to have a concentration of approximately 7MPm-3. Topographic and vegetative variables were measured using digital elevation and canopy height models, while hydraulic variables were calculated with an unsteady flow model in HEC-RAS. Sediment and microplastic accumulation were both found to be significantly influenced by terrain and hydraulic conditions. Sediment accumulation yielded an MLR model with an R2 of 0.72, with a confidence level between 97% - 99%, while the microplastic model yielded an R2 of 0.26 and with a confidence level between 95% - 98%. Differences in sediment and microplastics particle density resulted in hydraulic conditions being more influential on microplastic accumulation with an R2 81.5% greater than any its terrain components. This research identified floodplain accumulation process drivers which could help to guide future management decisions regarding sediment storage and monitoring microplastic accumulation.
  • A model generalization study in localizing indoor cows with cow localization (colo) dataset
    Das, Mautushi (Virginia Tech, 2024-07-10)
    Precision livestock farming increasingly relies on advanced object localization techniques to monitor livestock health and optimize resource management. In recent years, computer vision-based localization methods have been widely used for animal localization. However, certain challenges still make the task difficult, such as the scarcity of data for model fine-tuning and the inability to generalize models effectively. To address these challenges, we introduces COLO (COw LOcalization), a publicly available dataset comprising localization data for Jersey and Holstein cows under various lighting conditions and camera angles. We evaluate the performance and generalization capabilities of YOLOv8 and YOLOv9 model variants using this dataset. Our analysis assesses model robustness across different lighting and viewpoint configurations and explores the trade-off between model complexity, defined by the number of learnable parameters, and performance. Our findings indicate that camera viewpoint angle is the most critical factor for model training, surpassing the influence of lighting conditions. Higher model complexity does not necessarily guarantee better results; rather, performance is contingent on specific data and task requirements. For our dataset, medium complexity models generally outperformed both simpler and more complex models. Additionally, we evaluate the performance of fine-tuned models across various pre-trained weight initialization. The results demonstrate that as the amount of training samples increases, the advantage of using weight initialization diminishes. This suggests that for large datasets, it may not be necessary to invest extra effort in fine-tuning models with custom weight initialization. In summary, our study provides comprehensive insights for animal and dairy scientists to choose the optimal model for cow localization performance, considering factors such as lighting, camera angles, model parameters, dataset size, and different weight initialization criteria. These findings contribute to the field of precision livestock farming by enhancing the accuracy and efficiency of cow localization technology. The COLO dataset, introduced in this study, serves as a valuable resource for the research community, enabling further advancements in object detection models for precision livestock farming.
  • Continuous Monitoring of High Risk Disaster Areas by Applying Change Detection to Free Satellite Imagery
    Roush, Allison Granfield (Virginia Tech, 2024-06-11)
    Natural disasters can happen anywhere causing damage to land and infrastructure. When these disasters occur in remote areas without much human traffic, it may take a long time for someone to notice that an event has occurred and to respond to it. Response time and damages could be reduced if the area could be remotely monitored. Many satellites pass over the Earth everyday collecting valuable imagery data that is free to access. However, this data can be difficult to process and use in practical applications such as monitoring an area for changes. Existing programs that use satellite imagery to monitor areas for changes can cost a significant amount of money making it inaccessible to most people. In this paper, a software program is introduced to automatically retrieve, process, and analyze free satellite imagery data and notify the user of significant changes in their area of interest (AOI). First, a software program was developed to automatically download a package of satellite imagery data from Planet Labs that met certain requirements for AOI, date, and cloud cover. A second software program was developed to download this data from the Google Cloud Storage (GCS) space and compare a current image to the composite of previous images in order to detect a change. This program then creates a figure to display the current image, the previous image, the difference area, and a summary table of the difference metrics. This figure is saved and emailed to the user if the differences are greater than the set threshold. This program is also capable of running automatically in the background of a computer every time it is logged in. The success of the program in correctly identifying areas of change was tested in three locations using historical satellite image data. The software was successful in identifying areas of change and delivering this information to the user in an easy to understand summary figure. Overall, the software was able to utilize free satellite imagery to detect changes in disaster areas and deliver a summary report to a user to take action showing that this software could be used in the future as an easy way to monitor disaster areas.
  • Projecting Planning-Related Climate Impact Drivers for Appalachian Public Health Support
    Larsson, Natalie Anne (Virginia Tech, 2024-07-10)
    Climate change is impacting the intensity, duration, and frequency of climatic events. With climate change comes a multitude of adverse conditions, including extreme heat events, changes in disease patterns, and increased likelihood and frequency of natural disasters, including in places previously not exposed to such conditions. Human health has foundations in the environment; therefore, these adverse climatic conditions are directly linked to human health. Rural communities in Appalachia are likely to experience negative consequences of climate change more severely due to unique geomorphology and sociopolitical realities of the region. Non-governmental organizations (NGOs) throughout the Appalachian region are currently working to build resilience and prepare for potential adverse effects from climate change. To aid in this process, projections of future climate scenarios are needed to understand possible situations and adequately prepare. In partnership with Ohio University and West Virginia University, this study aims to characterize potential future climatic scenarios from publicly-available global climate models (GCMs) and prepare information to share with Appalachian communities. Climate model information for this analysis was obtained from NASA's Coupled Model Intercomparison Project (CMIP6). All code for data processing and analysis was prepared using the open-source R programming language to support reproducibility. To confirm that models can accurately simulate Appalachian climatic conditions, CMIP6 hindcast simulations for precipitation and maximum temperature were compared to observed weather records from NOAA. Climate models over and underestimated average precipitation values depending on location, while models consistently underestimated extreme precipitation values, simulated by total five-day precipitation. For temperature, climate models consistently underestimated average and extreme high temperature indicators. For Appalachian region projections, three towns of interest (one for each state involved in the study: Virginia, West Virginia, and Ohio) were selected based on current community resilience efforts. In these locations, mid-century (2040 – 2064) and end-of-century (2075 – 2099) projections for precipitation and temperature were summarized under a low emissions scenario and a high emissions scenario. Increases in precipitation and temperature were observed under average and extreme scenarios; these increases were noticeably more extreme under higher emissions scenarios. These trends are consistent with other studies and climate science consensus. When compared to hindcast values, observed average precipitation values were overestimated and underestimated, while observed extreme precipitation indices, average temperatures, and heat wave indices were underestimated by GCMs. Context with observed data is important to understanding model accuracy for the Appalachian region. GCMs are a useful tool to project potential future climate scenarios at specific locations in the Appalachian region, though model data is best used to communicate general trends rather than as inputs for other physical models.
  • On the implications of unsafe eBPF composition
    Somaraju, Sai Roop (Virginia Tech, 2024-06-10)
    In the era of Linux being omnipresent, the demand for dynamically extending kernel capabil- ities without requiring changes to kernel source code or loading kernel modules at runtime is increasing. This is driven by numerous use cases such as observability, security, and network- ing, which can be efficiently addressed at the system level, underscoring the importance of such extensions. Any extension requires programmers to possess high levels of skill and thor- ough testing to ensure complete safety. The eBPF subsystem in the Linux kernel addresses this challenge by allowing applications to enhance the kernel's capabilities at runtime, while ensuring stability and security. This guaranteed safety is facilitated by the verifier engine, which statically verifies BPF code. In this thesis, we identify that the verifier implicitly relies on safety assumptions about its runtime execution environment, which are not being upheld in certain scenarios. One such critical aspect of the execution environment is the availability of stack space for use while executing the BPF program. Specifically, we high- light this fundamental issue in certain configuration of the BPF runtime environment within the Linux kernel and how this unsafe composition allowed for kernel stack overflow, thus violating safety guarantees. To tackle this problem, we propose a stack switching approach to ensure stack safety and evaluate its effectiveness.
  • Impact of Attitudes on the Relationship between Psychological Symptoms and Help Seeking Behavior in a Black and Non-Black International Sample
    Jones, Sydney B. (Virginia Tech, 2024-04-08)
    Internationally, members of the African diaspora, (Black people), report higher rates of untreated mental illness than peers of other races. Research has suggested that symptoms associated with poor mental health such as clinical depression and anxiety are associated with negative evaluations of help seeking behaviors such as contacting mental health professionals for care. The current study sought to examine the impact of attitudes toward seeking mental health care on the causal relationship between symptoms of depression, anxiety or stress as measured by the DASS-21 to help seeking behaviors reported by participants. This study further examined the impact of racial identity on this relationship to highlight any discrepancies specific to Black people. This research is intended to help guide and improve outreach, access, and clinical approaches to treating Black people with mental illness. A total of 500 participants were recruited for this study via online surveying software. Participants were divided into two groupings of 250 Black participants and 250 Non-Black participants (N=500) to complete the survey. A moderated mediation analysis was conducted to examine the mediating effects of attitudes towards professional help seeking on the relationship between psychological symptomology and help seeking behaviors, as well as to examine any moderating effects that could be highlighted by racial differences. There was a significant direct relationship between symptoms and help seeking behaviors found with a significant partial mediating effect of participant attitudes on the direct relationship (R2= 0.1521, p=<0.000). Race was not found to be a significant moderator of this mediation (CI95%: -0.001 to 0.004), though race did moderate the direct relationship from symptoms to help seeking behaviors (β= 0.016, SE= 0.0025, t= 6.375, p= < 0.000).
  • BCC’ing AI: Using Modern Natural Language Processing to Detect Micro and Macro E-ggressions in Workplace Emails
    Cornett, Kelsi E. (Virginia Tech, 2024-05-24)
    Subtle offensive statements in workplace emails, which I term "Micro E-ggressions," can significantly impact the psychological safety and subsequent productivity of work environments despite their often-ambiguous intent. This thesis investigates the prevalence and nature of both micro and macro e-ggressions within workplace email communications, utilizing state-of-the-art natural language processing (NLP) techniques. Leveraging a large dataset of workplace emails, the study aims to detect and analyze these subtle offenses, exploring their themes and the contextual factors that facilitate their occurrence. The research identifies common types of micro e-ggressions, such as questioning competence and work ethic, and examines the responses to these offenses. Results indicate a high prevalence of offensive content in workplace emails and reveal distinct thematic elements that contribute to the perpetuation of workplace incivility. The findings underscore the potential for NLP tools to bridge gaps in awareness and sensitivity, ultimately contributing to more inclusive and respectful workplace cultures.
  • Connecting People In Motion
    Johnson, Graesen Elisabeth (Virginia Tech, 2024-06-12)
    Perception of movement within and between designed spaces starts with the uniquely human ability to relate ourselves to our surroundings, followed by a relationship to the sequential experience of our movement throughout. Architecture is simply a building without the life and movement of the people who use the design, yet individuals may experience and relate to the same design differently. Habitual routes and repetitive paths of movement dull our experience of these spaces while moving towards or within a new space can allow our perception to expand as we take in a new environment, creating excitement but also tension within us. At our center, there is a phenomenological connection between a preceding space and personal orientation with a future space, helping us understand the new space in relation to ourselves, no matter the mode of transportation for arrival. Transportation hubs are intersections of time, connecting people in motion and guiding both habitual and unfamiliar subjects along their continuous journey. Studying the movement within the Washington metropolitan area, the New Carrollton Station in Maryland perforates the Capital Beltway as a gateway to the region. This thesis aims to understand how people interact with path-connected spaces and connect each subject's mode of arrival, goal, and choice of movement between a newly designed station.
  • Competitive status of red spruce (Picea rubens) and Fraser fir (Abies fraseri) at ecotonal transitions in southern Appalachian sky islands
    Wetzel, Rose (Virginia Tech, 2024-07-05)
    Southern Appalachian spruce-fir sky islands are globally threatened, boreal relict forests where red spruce (Picea rubens) and Fraser fir (Abies fraseri) are dominant. Fraser fir dominates at the highest elevations with spruce-fir and spruce-dominated stands at middle elevations and hardwoods associating at lower elevations. A primary concern is encroachment of hardwoods upslope as climate change-driven milder temperatures and high precipitation confine spruce-fir forests to even higher elevations. We performed a dendrochronological analysis of growth rates in red spruce, Fraser fir, and competing hardwoods between cover types and slope aspects at six sky islands. We created linear models to test effects of aspect, cover type, and year on basal area growth measurements of red spruce, Fraser fir, and hardwoods to assess effects of competition. Growth rates were significantly affected by species, aspect, cover type, and year, and generally increased over time. Red spruce growth rates varied by combination of aspect and cover type but were greater than those of hardwoods on northern and southern aspects. Fraser fir growth rates were negative on southern-facing fir-dominated stands but increased in all other stands with the highest growth rates found in fir-dominated stands. The differences we report by cover type and aspect could help conservation practitioners prioritize treatment locations to improve climate resiliency.
  • Rapid FTIR analysis for respirable crystalline silica monitoring in coal mines using readily available sampling equipment
    Elie, Garek Christopher (Virginia Tech, 2024-07-01)
    In coal mines, workers can be exposed to respirable coal mine dust (RCMD) in conjunction with respirable crystalline silica (RCS). Overexposure can pose serious health risks, including development of coal workers' pneumoconiosis (CWP) (also known as "black lung"). CWP has the potential to progress to a more consequential form known as progressive massive fibrosis (PMF), for which a dramatic resurgence has been observed among US miners since the early 2000's. Recent rules promulgated by the Mine Safety and Health Administration (MSHA) have lowered the permissible exposure limit (PEL) of RCMD and RCS, but the nuances of dust monitoring are complicated. For RCMD, frequent monitoring is required using the continuous personal dust monitor (CPDM), which enables real time data—but the physical sample collected by the CPDM cannot currently be used for RCS analysis. For RCS monitoring, filter samples are still collected with the traditional coal mine dust personal sampling unit (CMDPSU)—but the standard RCS analysis must be done in a centralized lab and there is considerable lag time between sampling and data availability. To enable rapid RCS analysis of filter samples, NIOSH has developed a direct-on-filter (DOF) Fourier transform infrared (FTIR) spectroscopy method for use with CMDPSU filter samples. It can be performed in the field with a portable instrument. NIOSH has also developed a compatible software called the Field Analysis of Silica Tool (FAST), which simplifies processing of the FTIR spectral data to yield RCS mass results. While not allowed to demonstrate regulatory compliance with the RCS PEL, this method could be quite useful for routine non-regulatory monitoring (e.g., to support research or engineering studies). However, adoption of the method may hinge on a variety of factors such as costs, ease-of-use, and the usability and reliability of generated data. This thesis reports a field study designed to demonstrate how the DOF FTIR method (with FAST) might be used by mines with relatively low-cost, off-the-shelf sampling components for the CMDPSU. The field study also demonstrates how the percentage of RCS in RCMD (in addition to RCS mass) can be estimated by simply pairing a CPDM with the CMDPSU during sampling. Understanding RCS percentage may be important for a variety of research or engineering applications. While the DOF FTIR method can work well for CMDPSU samples, it is recognized that RCS analysis of CPDM samples would be ideal. However, the materials and construction of the filter assembly used by the CPDM is not conducive to DOF analysis. As part of an effort to develop a simple method for CPDM sample recovery, redeposition, and analysis by FTIR, the second study in this thesis focused on establishing the recovery procedure—and corrections to account for sample mass and RCS content attributed to any residue sourced from the CPDM filter assembly itself. Using blank CPDM filters and blank CPDM filters spiked with well characterized respirable dust, results show that the mass and RCS content of the CPDM residue may be quite small. Moreover, using field CPDM samples, results show that dust recovery can be quite high. Taken together, these are promising findings and suggest that a method for RCS analysis of CPDM samples is possible.
  • Robotic Pruning for Indoor Indeterminate Plants
    Srivastava, Chhayank (Virginia Tech, 2024-07-01)
    This thesis presents an innovative agricultural automation technique which focuses on addressing the significant perception challenges posed by occlusion within environments such as farms and greenhouses. Automated systems tasked with duties like pruning face considerable difficulties due to occlusion, complicating the accurate identification of plant features. To tackle these challenges, this work introduces a novel approach utilizing a LiDAR camera mounted on a robot arm, enhancing the system's ability to scan plants and dynamically adjust the arm's trajectory based on machine learning-derived segmentation. This adjustment significantly increases the detection area of plant features, improving identification accuracy and efficiency. Building on foreground isolation and instance segmentation, the thesis then presents an automated method for identifying optimal pruning points using best pose view images of indeterminate tomato plants. By integrating advanced image processing techniques, the proposed method ensures the pruning process by targeting branches with the highest leaf load. Experimental validation of the proposed method was conducted in a simulated environment, where it demonstrated substantially enhanced performance. In terms of pruning point identification, the method achieved impressive results with 94% precision, 90% recall, and 92% F1 score for foreground isolation. Furthermore, the segmentation of isolated images significantly outperformed non-isolated ones, with improvements exceeding 30%, 27%, and 30% in precision, recall, and F1 metrics, respectively. This validation also confirmed the method's effectiveness in accurately identifying pruning points, achieving a 67% accuracy rate when compared against manually identified pruning points. These results underscore the robustness and reliability of the approach in automating pruning processes in agricultural settings.
  • Determining habitat and biotic factors driving puma (Puma concolor) space use and underlying dynamic processes (colonization and extinction) over 20 years in protected and private areas throughout Belize, Central America
    McPhail, Darby K. (Virginia Tech, 2024-07-01)
    Despite being a top carnivore, there is relatively scant information on pumas (Puma concolor) in the neotropics especially compared to the more well-studied jaguar (Panthera onca). Understanding long-term puma distribution can affect land management decisions such as appropriate size of buffer zones around protected areas since pumas influence, and are influenced by, sympatric carnivore populations, lower trophic levels, and habitat. We used single-species, single-season and multi-season occupancy modeling to explore factors influencing distribution and persistence of pumas across the country of Belize. We used camera trapping data from 7 protected areas over 20 years with 2,198 camera stations covering ~5,000 km2. For both approaches, detection was mostly affected by distance to roads, enhanced vegetation index (EVI), and elevation, with variable directionality depending on site. In single season modeling, Occupancy increased at lower elevations and intermediate EVI in one site, and closer to water sources at another, while in multi-season modeling, intermediate EVI and canopy cover influenced occupancy. Biotic covariates were highly variable across sites and methods, but detection and occupancy were generally positively associated with prey, jaguar and ocelot trap rates, canopy cover, and elevation, while human trap rates negatively affected occupancy at one site. Colonization was positively affected by deer (Odocoileus virginianus and Mazama americana) trap rates while extinction had no supported covariates. Puma occupancy ranged from 0.41-0.96 in single season models and 0.55-0.90 in multi-season models across all site/years. Compared to other single-season studies, Belize generally had higher occupancy, even in areas of selective logging, however there are no other multi-season studies to compare. While sites with heavy human impacts had lowest occupancy, these areas are still used and likely serve as steppingstones between protected areas of higher occupancy. Such areas could be targets for protection to preserve landscape connectivity. Additionally, due to high occupancy and colonization across varying habitat and biotic factors the jaguar is likely an effective umbrella species for puma space use, however more analysis on other species is needed to ensure efficiency for more than just pumas.
  • Something that Hasn't Happened Yet
    Wilson, Christopher (Virginia Tech, 2019-07-02)
    Something that Hasn't Happened Yet is a collection of poems wherein the speaker traverses the world of family and relationships in an absurdist/meta-modern narrative. The collection also explores the form of poetry itself as well as creative nonfiction, recipes, and flash fiction as well. It is an attempt to assemble meaning through humor and the process of writing itself.
  • Someplace Else
    Jernegan, Leslie Erin (Virginia Tech, 2019-07-02)
    The novel Someplace Else scrutinizes spaces begging for examination—places of asphyxiation, of undiscussed power structures and violence—that do nothing to prepare those living within them to be their examiners. Through the lens of Lumi—a small-town Wisconsin adolescent on the verge of womanhood—the novel examines how childhood innocence is exemplified and threatened by the homes in which females are raised and raising themselves. Someplace Else serves as Lumi's avenue for figuring out how to put to words what exactly it is she is coming to understand, including her relationship with her hometown, how this space has affected her mother and sister, and how this space has affected these women's relationships with one another; through story, Lumi is deciphering ways to speak, to talk about her world and perhaps find a way out.
  • Matryoshka
    Cohen, Tali Sharon (Virginia Tech, 2019-07-02)
    Matryoshka is a poetry collection that inhabits the space between danger and desire. The poems are largely voice-driven and confessional, sprouting from a speaker who is somehow ruthlessly honest and deceptively evasive at the same time. She covets the domestic only to set it on fire. She runs for comfort and greets the comfortable with a knife. In the beginning of the collection, we see a speaker navigating her relationships with others. The poems in section one are seeped with intense longing and physical desire. In section two, we see the speaker turn her gaze inward and are met with a raw exploration of the self; a revelry of bad decision making, self-deception, and complicated sexuality. The collection leaves the reader curious and comfort-seeking.
  • The NGOization of Indian Agricultural Development and Implications for the Agrarian Question
    Hammond, Erin Elizabeth (Virginia Tech, 2024-06-27)
    India failed to answer the agrarian question after independence by not undertaking expansive land reforms and rural redistribution of resources and wealth. Instead, India followed the national bourgeois path of development, liberalizing agricultural production systems based on rural bourgeoisie and foreign interests. This path of development has led to unequal rural development and the NGOization of agricultural development. For agrarian and peasant producers in India, the liberal and privatized NGO development path of the Indian agriculture sector has had significant implications for their social, political, and economic well-being. The role of the private, for-profit NGO in the liberal, capitalist agriculture production system has not been given as much attention as the role of the Indian and foreign governments, international institutions, and transnational corporations. This thesis argues that private, for-profit non-governmental development organizations working in rural India reproduce imperial structures of foreign dependency and increases the subsumption of peasant and agrarian producers. NGOization conceals the global power structures at play within Indian agricultural production and can impede upon alternative solutions to the Indian agrarian question by appropriating local thought leaders, grassroots movements' narratives, and Indigenous knowledge, which further perpetuates imperial and colonial structures of rural communities and leads to the de-depeasantization of rural production systems.
  • A Machine-Learning Based Approach to Predicting Waterborne Disease Outbreaks Caused by Hurricanes
    Mansky, Christopher Immanuel (Virginia Tech, 2024-06-27)
    Climate change is increasing the frequency and intensity of (extra-) tropical cyclones including hurricanes and winter storms worldwide. Waterborne diseases, resulting from flood-related impacts, affect public health and are of major concern for society. Previous research studies have highlighted a statistically significant linear correlation between waterborne diseases and climate variables, especially precipitation and temperature. However, to the best of our knowledge, no studies have explored nonlinear models (e.g., machine learning) to predict waterborne disease outbreaks in the aftermath of hurricanes and winter storms. Here, we aim at predicting waterborne disease counts as well as disease outbreaks using historic climate demographic, and public health data of Florida, U.S. that date back to 1992. For this, we first predicted diseases in aggregated coastal counties using multiple linear (MLR) and random forest regression (RFR) models. Then, we developed a binary random forest classifier (RFC) model to predict waterborne disease outbreaks (e.g., 0: no outbreak and 1: outbreak). Results of this study showed that the highest coefficient of determination (R2) for the MLR model was 0.65 for two aggregated county groups, namely St. Johns-Duval-Nassau and Sarasota-Charlotte-Lee. The RFR model showed the highest R2 of 0.69 for the county group Sarasota-Charlotte-Lee. The highest Root Mean Square Error (RMSE) was found for the county group Miami Dade-Broward- Palm Beach with a value of 15 and 16 people for both the MLR and RFR models. St. Johns-Duval-Nassau and Sarasota-Charlotte-Lee groups achieved the highest Kling-Gupta Efficiency (KGE) of 0.76 for the MLR model. Sarasota-Charlotte-Lee also performed the best in terms of KGE for the RFR model with a score of 0.69. On the other hand, the binary RFC model for Pinellas-Hillsborough-Manatee achieved a model's accuracy of 0.93 and f1-score of 0.48. We anticipate that the models' performance can substantially be improved with access to higher spatial resolution climate data as well as longer demographic and public health records. Nevertheless, we here provide a solid methodology that can inform local authorities about imminent public health impacts and mitigate negative effects on society, economy, and environment.