113588 research outputs found
Sort by
Application of Machine Learning Methods for Predicting the Factor of Safety in Rock Slopes
No full textThe mining industry is driven by an increasing worldwide demand for minerals that necessitatesthe continuous improvement of safe and cost-effective excavation operations. Slope stability is a primary concern for every mine to avoid significant safety issues and economic loss. Deterministic and stochastic methods for estimating the Factor of Safety (FOS) have been widely implemented and improved for several decades to overcome the uncertainties and interactions among geotechnical parameters in rock slopes, but these methods are limited in analyzing the nonlinear input/output relationships between a large number of variables and in their application to different geotechnical conditions. Recent progress in machine learning (ML) can serve as an alternative method for forecasting FOS by modeling the relationships between inputs and outputs, providing a more reliable estimation of slope stability. In this research, a literature review of ML applications for slope stability assessment was conducted to gain a comprehensive understanding of the topic. In addition, four ML models—Gaussian Process Regressor (GPR), Support Vector Regressor (SVR), Random Forest (RF), and a hybrid Genetic Algorithm–Multi-Layer Perceptron (GA-MLP)—were implemented and compared on two real-world datasets. The datasets have similar geotechnical parameters from the excavated slopes and use the measured FOS as the output label. The first dataset was collected from a highway excavation site in China, and the second from a mining operation in Peru. The geotechnical parameters included in the datasets are slope height, slope angle, unit weight, cohesion, and friction angle. Each dataset was split into training, validation, and testing sets. The accuracy and consistency of the four ML models on both datasets were subsequently determined. The SVR model showed the best accuracy on the mining dataset, while the GPR model was superior on the highway dataset and demonstrated more consistency across both
Evaluating Embedding Techniques for Emotion Classification in Machine Learning
No full textThis thesis investigates the effectiveness of different embedding techniques for emotion classification in natural language text. Using the GoEmotions dataset, we evaluate multiple embeddings including contextual and dimensionality-reduced variants across various classification models such as BERT, Bi-GRU, and logistic regression. The study focuses on a multi-label emotion detection task, where each emotion is treated as an independent binary classification problem. Performance is assessed using standard evaluation metrics, and emotion-specific threshold selection is employed for improved prediction. Our findings highlight the impact of embedding choice on classification accuracy and provide insights for building more robust emotion-aware NLP systems
Understanding Public Preferences for Non-Native Species Management and Wildfire Risk Mitigation in Arizona
No full textWildfires in Arizona are increasing in frequency and severity, often fueled by invasive plant species such as buffelgrass. This thesis explores public preferences for wildfire mitigation strategies, particularly invasive species management, using a Best-Worst Scaling (BWS) survey supported by Conditional Logit (CLM) and Random Effects Logit (REL) models. The study focuses on high-risk communities near the Tonto National Forest, using a stakeholder-informed survey design. Results show that residents prioritize tangible safety outcomes over institutional or financial considerations—especially “Protection of Homes,” “High to Medium Risk Reduction,” and “Mechanical Thinning.” These attributes consistently ranked highest in the BWS and were statistically significant in the CLM model. In contrast, respondents expressed strong opposition to utility tax increases, regardless of magnitude. The CLM results confirmed that price was a significant negative predictor of preference, alongside positive preferences for safety-focused attributes. However, the REL model, used to simulate actual voting behavior under cost constraints, revealed that price alone was statistically significant. Non-price attributes, such as the type of administering agency, prevention method, or prioritization goal, had no measurable influence on voting outcomes once cost was introduced. This suggests a divergence between symbolic support for mitigation strategies and willingness to pay—likely driven by the public goods nature of wildfire protection, in which individuals may support collective benefits in principle but prefer to free ride when financial contributions are required. This work suggests that policymakers should consider designing low-cost, high-impact programs and explore non-fee-based or collectively funded mechanisms to align wildfire mitigation strategies with public preferences. Doing so may improve political feasibility and reduce resistance among cost-sensitive or lower-income households
BUILDING A WEB GIS APPLICATION FOR LOCATING PARKS AND AMENITIES IN ROSEVILLE, CALIFORNIA
No full textParks are an important part of the community. They provide a place for the public to gather for events, play sports, enjoy the outdoors, and catch up with neighbors. Roseville, California, is a growing city that is continually developing new parks, which can make it difficult for the public to be aware of all the parks and amenities the city has to offer. City parks are often an underutilized public resource. Lack of park awareness is one of the main reasons cited for not utilizing city parks. In Roseville, the public may visit the city’s Parks and Places webpage to view a map of city parks. To find particular park amenities, the public may visit 88 individual park webpages to view each of the static lists of available park amenities for each park. Through this project, an interactive web application is built to eliminate the time-consuming process when searching for park amenities and to provide awareness of the parks and park amenities offered throughout the city of Roseville. ArcGIS Pro, ArcGIS Online, and ArcGIS Experience Builder are used to build an application that allows the user to search for parks by name, map location, or by a desired combination of park amenities. This application will serve as a prototype web application that will allow the public to search Roseville parks and park amenities in a user-friendly way, bringing awareness to all parks and park amenities offered in Roseville.This item is part of the MS-GIST Master's Reports collection. For more information about items in this collection, please contact the UA Campus Repository at [email protected]
Flood Impact Analysis of Typhoon Yagi in Vietnam
No full textOn September 7, 2024, Typhoon Yagi struck Northern Vietnam, severely affecting the coastal city of Hai Phong and neighboring Quang Ninh Province. The typhoon caused extensive flooding, significant infrastructure damage, and substantial economic and human losses. This research integrates remote sensing and GIS-based spatial analysis to evaluate the flood impacts and predict flood-prone areas. Sentinel-1 satellite data, processed using the Sentinel Application Platform (SNAP), was employed to extract flood-affected regions through pre- and post-crisis imagery differencing. Flood extent was identified by subtracting pre-crisis from post-crisis imagery, isolating inundated areas. Additionally, a predictive flood risk analysis was conducted using the weighted sum tool in ArcGIS Pro, incorporating seven key variables: DEM, terrain slope, drainage density, LULC derived from Sentinel-2 data, monthly rainfall data for September 2024 from PERSIANN-CCS, distance to roads, and distance to rivers. Each raster variable was reclassified into a standardized scale ranging from 1 (very low risk) to 5 (very high risk) to ensure comparability, followed by the assignment of weighted contributions to each variable based on its influence on flooding. The resulting analysis produced a flood risk map for the study area, highlighting the interplay of topography, hydrology, and infrastructure in flood dynamics. These findings offer critical insights for flood risk assessment, disaster response, and the development of mitigation strategies tailored to Hai Phong and Quang Ninh.This item is part of the MS-GIST Master's Reports collection. For more information about items in this collection, please contact the UA Campus Repository at [email protected]
A Suitability Analysis of Dona Ana and El Paso Phase Settlement Patterns in the Jornada Mogollon Region of New Mexico
No full textThe goal of this project is to use raster-based GIS analytical methods, specifically suitability analysis, to identify and model landscape characteristics associated with archaeological site placement in southeastern New Mexico. This analysis focuses particularly on the environmental variables influencing the transition from mobile hunter-gatherer lifeways in the Jornada Mogollon/Doña Ana phase to increasingly sedentary agricultural settlements typical of the later El Paso phase. By integrating factors such as proximity to streams, slope, aspect, precipitation, soils, and landcover into a suitability model, this research aims to better understand how these variables influenced settlement decisions.
During the Late Formative (AD 1000 – 1450), Jornada Mogollon populations adopted a mixed subsistence strategy characterized primarily by horticulture focused on indigenous plant species (e.g., chenopods), supplemented by maize agriculture. This adaptive approach would have had important implications for where communities established their settlements, likely emphasizing areas suitable for small-scale agriculture and resource-rich environments. Identifying landscape characteristics associated with semi-sedentary to sedentary residential sites will help clarify the environmental criteria guiding site selection and reveal broader patterns in prehistoric settlement decisions. This research offers insights into how environmental constraints shaped cultural adaptations and landscape utilization in the Jornada Mogollon region.This item is part of the MS-GIST Master's Reports collection. For more information about items in this collection, please contact the UA Campus Repository at [email protected]
Expanding the Primary Care Scholarship: Addressing Physician Shortages by Including Emergency Medicine in Rural and Underserved Arizona
No full textArizona is facing an escalating Primary Care Physician shortage, with rural and underserved communities experiencing the most severe impacts. Despite their critical role in providing frontline care, Emergency Medicine physicians are currently excluded from the University of Arizona Primary Care Scholarship—a program designed to incentivize healthcare providers to work in Health Professional Shortage Areas. This thesis advocates for expanding scholarship eligibility to include Emergency Medicine physicians, highlighting their in-practice role as primary care providers in high-need settings. Through a comprehensive analysis, this thesis redefines what constitutes primary care in today’s healthcare landscape and examines the inconsistency in specialty classification across institutions, and further expansion of the scholarship criteria to accommodate and include Emergency Medicine. For instance, while the Association of American Medical Colleges does not classify General Surgery or Psychiatry as primary care, the University of Arizona includes them as eligible under the scholarship—raising questions about the criteria used. Meanwhile, the University of Arizona’s own Primary Care Physician Workforce Report recognizes Emergency Medicine as a primary care specialty and supports rural emergency rotations for Emergency Medicine residents. The thesis also presents original workforce data showing that Emergency Medicine ranks second behind Family Medicine in the number of physicians practicing in rural areas, surpassing Internal Medicine, Pediatrics, and General Surgery. It documents a growing national trend of patients turning to emergency departments for non-emergent, primary care-level services, and details how the Emergency Medicine physician shortage forces Family Medicine doctors to split time between clinics and emergency departments—ultimately reducing primary care productivity. Curriculum analysis from the University of Arizona South Campus Emergency Medicine residency program reveals that Emergency Medicine residents, especially those on the Rural/Border/Global Health Track, receive substantial primary care-relevant training. Additionally, this report explores how expanded training, integrated systems, and financial incentives—like loan repayment and scholarships—are crucial to supporting Emergency Medicine physicians' dual role. By making Emergency Medicine physicians eligible for the University of Arizona Primary Care Scholarship, Arizona can better align its workforce strategies with on-the-ground realities, ultimately strengthening care access and equity in underserved regions.Replaced file with corrected version 3/20/2026 per Graduate College, Kimberl
Changes in Federal Law and Policy: Effects on Cultural Resources in Organ Pipe Cactus National Monument
No full textThis study examines how changes and waivers to federal regulations affect cultural resource management (CRM) professionals and the cultural resources they protect, focusing on Organ Pipe Cactus National Monument (ORPI) and the southern Arizona border region. Through document review and interviews with CRM practitioners, the research explores the impacts of the 2020 amendments to the National Environmental Policy Act (NEPA) and the waivers issued under Section 102 of the REAL ID Act of 2005 for border wall construction. The findings reveal that while the 2020 NEPA changes had limited immediate effects on CRM professionals and cultural resources in ORPI, the use of law waivers under the REAL ID Act led to significant adverse impacts. The waivers removed legal requirements for environmental and cultural resource review, leading to insufficient impact prediction, reduced mitigation, and greater harm to resources and heritage sites, particularly at Quitobaquito Springs and along traditional Indigenous routes. The study underscores the importance of regulatory processes, agency communication, and professional involvement for effective cultural resource stewardship. It concludes that although regulatory streamlining may not greatly alter CRM practices in low-project areas like ORPI, removing legal safeguards through waivers poses substantial risks to both cultural resources and the professionals responsible for their protection
Alterations in Hippocampal-Prefrontal Navigation Circuits and Behaviors With Age
No full textAge-related declines in episodic and working memory processes have been well established in mammals. Spatial navigation, of which mnemonic processes are a component, requires the integration of inputs from multiple brain regions, each individually affected by aging, making it difficult to identify the mechanisms underlying these age-related changes. Two key regions thought to be involved in navigation are the hippocampus and prefrontal cortex. One hypothesis is that the hippocampus integrates environmental and self-motion cues to build spatial representations, while the medial prefrontal cortex (mPFC) draws on those representations to guide behaviors. These regions are connected via direct projection from the hippocampus to the prefrontal cortex along with other indirect bidirectional projections between them. Neural activity between both of these regions is thought to be coordinated through local field potential oscillations that facilitate the transfer of information between them. This thesis provides insights into how aging specifically affects hippocampal-prefrontal cortex (HC-PFC) circuits that in turn influences navigation performance. In spatial navigation tasks, despite learning optimal spatial allocentric strategies, aged rats fail to deploy them consistently, resulting is slower spatial learning compared to their younger counterparts. They also perform worse on spatial working-memory tasks that require HC–mPFC coordination. In young rats, better performance is accompanied by stronger oscillatory synchrony in the theta band (6-12Hz) between the two regions, but this relationship weakens in aged rats that have overall reduced theta coherence. Circuit-level experiments in anesthetized rats reveal a selective attenuation of ventral hippocampal input to the infralimbic cortex, that might influencereduced coordination between regions. Additionally, specific impairments to the infralimbic region, may shift the prelimbic–infralimbic balance resulting in increased perseverative or context-inappropriate responses with age. Together, these findings demonstrate that aging correlates with region-specific alterations in HC-PFC circuits, with ventral hippocampal-infralimbic connectivity being particularly vulnerable. This selective impairment shifts the balance within mPFC circuits, potentially leading to alterations in behavioral strategies utilized in specific spatial contexts as observed in aged animals, ultimately disrupting the neural computations necessary for effective spatial navigation.Release after 11/25/202
Non-Destructive Evaluation of Sustainable Cementitious Materials Using Ultrasonic Techniques
No full textThis dissertation investigates sustainable cementitious materials through advanced non-destructive ultrasonic evaluation, integrating two complementary strategies: waste rubber incorporation and accelerated carbonation curing. The first phase examines recycled rubber fiber mortars (0–20% by volume) to valorize tire waste and enhance durability. Frequency-domain ultrasonic analysis, using the Sideband Peak Count - Index (SPC-I), revealed high sensitivity to microstructural damage and strong correlation with compressive strength. The normalized SPC-I increased to 1.42 at 10% rubber, corresponding to peak nonlinearity and lowest strength, followed by partial recovery at higher contents. Rubber-modified mortars demonstrated improved ductility and reduced surface heating, suggesting suitability for non-structural and pavement applications. The second phase explores accelerated carbonation curing as a carbon capture and performance-enhancement method. Mortar specimens exposed to controlled CO? conditions (1–28 days) were characterized using Ultrasonic Pulse Velocity (UPV), nonlinear ultrasonics (SPC-I), strength testing, and gravimetric CO? uptake. Nonlinear ultrasound effectively captured microcrack closure and matrix densification, where prolonged carbonation reduced SPC-I values and improved compressive strength by up to 38.9%. CO? uptake reached 2.62%, confirming successful sequestration and linking acoustic response to chemical transformation. Collectively, findings establish nonlinear ultrasonic parameters, particularly SPC-I, as robust indicators of microstructural evolution in sustainable mortars. This research advances performance-based, low-carbon design of cementitious materials by demonstrating how non-invasive acoustics can quantify both environmental benefit and mechanical performance.Release after 07/13/202