-
Prospective Students
-
Faculty and Staff
-
Current Students
-
Alumni and Friends
Genomic epidemiology is the use of genetic data to characterize and explain disease occurrence and transmission. Application of these methods to malaria has already yielded substantial benefits, such as identification and surveillance of drug resistance genotypes. However, the potential for genomic epidemiology to accelerate progress towards malaria eradication is far from fully realized. This dissertation demonstrates new applications of genomic information to questions that are impossible to address with conventional epidemiological data. First, the value of correlating genetic and environmental distances to understand the drivers of Plasmodium falciparum transmission is showcased with microsatellite data from 44 sites in Western Kenya. Second, the design and validation of a new panel of genetic markers, microhaplotypes with multiple SNPs on each short read, is presented for P. vivax, enabling sensitive, scalable characterization of within-host diversity in multi-strain infections. Finally, a similar panel of microhaplotype markers for P. falciparum is applied to samples from eight countries throughout Africa, yielding insights into continent scale transmission dynamics. The analysis of environmental drivers revealed the Winam Gulf of Lake Victoria as a barrier to malaria transmission, a conclusion that would be impossible to reach rigorously without this novel methodology. The new P. vivax panel yielded quality sequences and detected expected patterns of genetic relatedness, indicating this tool is ready for broad application. The P. falciparum microhaplotype analysis identified subtle patterns of genetic relatedness and surprisingly little relationship between within-host diversity and incidence, highlighting the potential of these markers but also a need for future work on the interpretation of the resulting data. This dissertation expands the scope of questions about malaria epidemiology that can be answered with genomic data and argues that routine application of these methods could accelerate progress towards malaria eradication.
This dissertation addresses critical aspects of traffic safety, focusing on novel approaches for weather-related crash prediction—a significant concern in the transportation field. It is divided into three interconnected studies: geospatial risk mapping of weather-related crashes, addressing data imbalanced in machine learning for weather-related crash severity analysis, and analytics for future weather-related crash prediction. In the first study, the dissertation advances a novel approach to hotspot mapping by developing a spatio-temporal cube that incorporates both the spatial and temporal dimensions of crash data, providing a dynamic and comprehensive analysis of crash hotspots. In the second study, the dissertation tackles the challenge of imbalanced data, which can bias machine learning model outputs, making them less adept at predicting crash severity. By extending methods such as Synthetic Minority Over-sampling Technique (SMOTE) and Adaptive Synthetic Sampling (ADASYN), the dissertation evaluates the effectiveness of these methods in datasets with a prevalence of nominal predictors, aiming to enhance the predictive accuracy of machine learning models for crash severity. Lastly, the dissertation proposes the use of Spatially Ensembled ConvLSTM algorithm for predicting a weather-related traffic crash. This approach aims address the limitations of traditional predictive models by leveraging the ability of LSTMs to retain relevant information over extended time frames across different heterogenous spaces. The proposed technique was compared with existing methods to test if it outperform conventional predictive models and the standard ConvLSTM in accuracy.
Racially minoritized students (RMS) face substantial disparities in college persistence and completion rates (Museus & Saelua, 2017). In particular, Black student enrollment at public two-year or community colleges has declined significantly, dipping below 13% in 2020, while for-profit institutions have maintained enrollment of Black students at roughly 28% over a 10-year period (AACC, 2023). Because community colleges have a reputation for being low-cost, high-quality institutions with more than 60 % of its graduates free of student loan debt (AACC, 2019), proper attention must be given and action taken to identify and address the needs of RMS in the community college settings to increase persistence and graduation rates. Sociological research on community colleges highlights the stratified tension between the increased provision of access as open-door institutions against low rates of successful completion (Schudde & Goldrick-Rab, 2014). While culturally relevant education practices have been most successfully implemented in the K-12 space (Ladson-Billings, 1995), an amplified call goes out to responsible community college leaders for the creation of culturally relevant campus environments. Using the culturally relevant leadership practices framework (Jones et al., 2016), this cross-case study explores the roles and practices of presidents and executive leaders within the context of their community colleges to determine how they create spaces for Black student achievement.
This dissertation explores the impact of student organizations on the persistence of women in STEM programs at the associate degree level. The findings reveal that participation in SkillsUSA provides students with valuable opportunities for hands-on learning, skill development, and career exploration, all of which contribute to the persistence of the participants in their chosen STEM related field. SkillsUSA offers a range of activities, including competitive events, leadership development, and community service projects, that foster collaboration, communication, and problem-solving skills among students (Maldonado & Jaeger, 2021; Threeton & Pellock, 2016). SkillsUSA can serve as a bridge between classroom instruction and real-world application, allowing students to apply their knowledge in authentic settings and gain practical experience in their chosen fields. Key themes that emerged from the data include the importance of mentorship, peer support, and extracurricular student engagement in shaping student’s academic and career trajectories. The participants of this study expressed gratitude for the guidance and encouragement provided by their SkillsUSA advisors and mentors, as well as the importance of the sense of camaraderie they developed with other women who were working to pursue a STEM career.
Labor dystocia, a term used to describe slowly progressing labor, is the most common reason for cesarean delivery. Despite global efforts to establish improved practice guidelines over the past decade, there is significant debate in the literature about how to manage labor dystocia when it occurs. The present study aims to illuminate 1) the decision-making processes surrounding labor dystocia, which previous literature suggests are complex and involve multiple stakeholders, and 2) the factors clinicians consider as part of these decisions that may contribute to whether a cesarean delivery is ultimately performed. These questions were approached qualitatively using informed constructivist grounded theory methodology. Informants were obstetricians, family medicine physicians, midwives, and labor and delivery nurses in current practice in metropolitan North Carolina hospitals. The primary researcher conducted semi-structured interviews that included a graphic elicitation diagramming exercise and collected sociodemographic data via an online survey. Several methodological strategies bolstered the study’s rigor and trustworthiness. The data revealed four common pathways through which decisions are made in the context of labor dystocia. Additionally, a Social-Ecological Model of Intrapartum Decision-Making is proposed that represents influential factors on decision-making processes at the level of the individual, patient-clinician, immediate social context, care team, maternity center/hospital setting, and broader macrosystem. Findings advance our understanding of how decisions are reached during a uniquely challenging medical experience and may lead to improvements in equitable, high-quality, labor and delivery care.
Trigger-action Internet of Things (IoT) platforms allow IoT devices to create a chain of interactions to automate network tasks by leveraging functional dependencies between IoT event conditions and actions. When network devices notify their cyber states to the IoT hub by reporting event conditions, the hub utilizes this chain to invoke actions in corresponding IoT devices dictated by user-defined rules. Adversaries exploit this scenario to implement remote injection attacks by maliciously reporting fake event conditions to the hub to force it to command target IoT devices to perform invalid actions violating rule integrity. Security mechanisms in the existing literature either require complete visibility over network events to provide an effective defense against dynamic injection attacks or do not offer real-time security.
In this dissertation, we present three security systems to fill this gap in the literature: 1) IoTMonitor, a Hidden Markov Model (HMM) based security analysis system that extracts optimized attack paths and discovers frequently exploited nodes in the network; 2) IoTWarden, a Deep Reinforcement Learning (DRL) based real-time defense system that allows a defense agent to learn attack behavior by observing the network environment and design an optimal defense policy to counter attacker's actions at runtime, maximizing overall security rewards; 3) IoTHaven, A POMDP-based online defense system to discern optimal defense policy for the partially observable IoT networks.
In this work, the in vitro characterization profiles of delivery vehicles, specifically polyamidoamine (PAMAM) dendrimers, are assessed to investigate their impact on pre-established immune responses to immunostimulatory and immunoquiescent nucleic acid nanoparticles (NANPs). Isolated human peripheral blood mononuclear cells (PBMCs) were used as the universal model system for these investigations, providing detailed understanding of the impact delivery vehicles play on NANP recognition. Additionally, to further identify mechanisms of immune recognition of these novel formulations, several engineered reporter cell lines were employed to understand the involvement of pattern recognition receptors, relevant to nucleic acid detections in human cells.
Furthermore, we explore the design and in vitro assessment of conditionally activated reconfigurable nucleic acid nanoparticles (recNANPs). By further investigating dynamic recNANPs and their interactions with delivery vehicles and the immune system, we aim to gain deeper insights into these systems. This innovative platform will enable the development of refined design principles for therapeutic systems incorporating NANPs, allowing for the creation of more precise and optimized options.
This dissertation enhances the Extended Unified Theory of Acceptance and Use of Technology (UTAUT2) by integrating information privacy concerns and examining their influence on adopting web-based healthcare portals. Through a survey of 298 U.S. residents using healthcare technologies, the study investigates the interplay between UTAUT2 predictors—Performance Expectancy, Effort Expectancy, Facilitating Conditions, Habit, Social Influence, and Hedonic Motivation—and the intention to use these technologies while assessing how privacy concerns modulate these relationships. Regression analysis highlights the positive impact of Performance Expectancy, Effort Expectancy, and Habit on adoption intent, with privacy concerns significantly moderating the relationship between Effort Expectancy and usage intention.
The research enriches the UTAUT2 model by showcasing the pivotal role of privacy concerns, thus advancing theoretical understanding and enhancing model predictability in the context of healthcare technology. Practically, it offers insights for practitioners and policymakers on addressing privacy concerns to improve technology adoption. This synthesis of privacy concerns within the technology acceptance framework paves the way for targeted strategies to increase the uptake of healthcare technologies, marking a significant contribution to both academic discourse and practical application.
This dissertation enhances the Extended Unified Theory of Acceptance and Use of Technology (UTAUT2) by integrating information privacy concerns, examining their influence on the adoption of web-based healthcare portals. Through a survey of 298 U.S. residents using healthcare technologies, the study investigates the interplay between UTAUT2 predictors—Performance Expectancy, Effort Expectancy, Facilitating Conditions, Habit, Social Influence, and Hedonic Motivation—and the intention to use these technologies, while assessing how privacy concerns modulate these relationships. Regression analysis highlights the positive impact of Performance Expectancy, Effort Expectancy, and Habit on adoption intent, with privacy concerns significantly moderating the relationship between Effort Expectancy and usage intention.
The research enriches the UTAUT2 model by showcasing the pivotal role of privacy concerns, thus advancing theoretical understanding and enhancing model predictability in the context of healthcare technology. Practically, it offers insights for practitioners and policymakers on addressing privacy concerns to improve technology adoption. This synthesis of privacy concerns within the technology acceptance framework paves the way for targeted strategies to increase the uptake of healthcare technologies, marking a significant contribution to both academic discourse and practical application in healthcare technology management.
This dissertation explores the intersection of Geographic Information Science (GIScience) and Artificial Intelligence (AI), specifically focusing on the enhancement of 3D deep learning models by spatial principles for understanding 3D geospatial data. With the rapid advancement in geospatial technologies and the proliferation of 3D data acquisition methods, there is a growing necessity to improve the capability of AI models to interpret complex 3D geospatial data effectively. This work seeks to leverage spatial principles, particularly spatial autocorrelation, to address the challenges pertaining to 3D geospatial object detection.
The research is structured around three pivotal questions: the utility of spatial autocorrelation features for understanding 3D geospatial data, the approach to derive content-adaptive spatial autocorrelation features, and the enhancement of post-processing in the task of 3D geospatial object detection. Through a series of experiments and model developments, this dissertation demonstrates that incorporating spatial autocorrelation features, such as semivariance, significantly enhances the performance of 3D deep learning models in geospatial object detection. A novel spatial autocorrelation encoder is introduced, integrating spatial contextual features into the 3D deep learning workflow and thereby improving accuracy in detecting objects within complex urban and natural environments. Further, the dissertation delves into the challenges brought by data partitioning and sampling in large-scale 3D point clouds, as evidenced in the DeepHyd project focusing on the detection of hydraulic structures (i.e., bridge and its components). The findings highlight the critical role of spatial dependency patterns in optimizing object detection accuracy and pave the way for future improvement of the 3D deep learning frameworks.
