Electronic Thesis or Dissertation

This dissertation focuses on the development of data-driven and physics-based modeling for two distinct significant structural engineering applications: time-varying response variables estimation and unwanted lateral vibration control. In the first part, I propose a machine learning (ML)-based surrogate modeling to directly predict dynamic responses over an entire mechanical system during operations. Any mechanical system design, as well as structural health monitoring systems, require transient vibration analysis. However, traditional methods and modeling calculations are time- and resource-consuming. The use of ML approaches is particularly promising in scientific and engineering challenges containing processes that are not completely understood, or where it is computationally infeasible to run numerical or analytical models at desired resolutions in space and time. In this research, an ML-based surrogate for the FEA approach is developed to forecast the time-varying response, i.e., displacement of a two-dimensional truss structure. Various ML regression algorithms including decision trees and deep neural networks are developed to predict movement over a truss structure, and their efficiencies are investigated. ML algorithms have been combined with FEA in preliminary attempts to address issues in static mechanical systems.

Many studies on shark swimming have examined kinematic variables along straight tracks or under controlled flow speeds in flumes, but there is less known about unsteady swimming during maneuvering or feeding. Sharks may adjust their speed, undulatory kinematics, or body curvature to accommodate different actions. This study quantified variations in kinematics during straight swimming, maneuvering, and feeding in scalloped hammerhead sharks (Sphyrna lewini). I obtained video of three juvenile scalloped hammerheads, developed an ethogram assessing three behavioral categories, and tracked points along the body’s midline. I found that velocity was lower during feeding compared to maneuvering and straight swimming, while body curvature increased during feeding turns but decreased with increasing velocity. These data will provide insight into kinematic variations in hammerhead sharks across ontogeny and among behaviors, ultimately expanding on the relationship between form and function. This also provides context for varying behaviors and trends within the movement ecology paradigm.

This thesis analyzes cranial modification from various sites and locations within coastal Ecuador. This research aims to identify the various types of tabular cranial modification and understand the methods used to classify each subtype of tabular modification. From this, I discussed the different types of modification and then used that information to contrast between North American and South American bioarchaeological methods of classifying cranial modification. Additionally, I reconstructed the biological profiles of some of the crania. The importance of this research is to introduce a method of identifying cranial modification that has been previously used in South American bioarchaeology to North American bioarchaeology. Furthermore, information on cranial modification regarding the coastal populations of Ecuador is lacking. The data in this thesis contributes a significant amount of knowledge about this practice, allowing this project to provide new information to the field of anthropology and the country of Ecuador.

Wall-bounded turbulent flows are pervasive in numerous physics and engineering applications. Such flows tend to have a strong impact on the design of ships, airplanes and rockets, industrial chemical mixing, wind and hydrokinetic energy, utility infrastructure and innumerable other fields. Understanding and controlling wall bounded turbulence has been a long-pursued endeavor yielding plentiful scientific and engineering discoveries, but there is much that remains unexplained from a fundamental viewpoint. One unexplained phenomenon is the formation and impact of coherent structures like the ejections of slow near-wall fluid into faster moving ow which have been shown to correlate with increases in friction drag. This thesis focuses on recognizing and regulating organized structures within wall-bounded turbulent flows using a variety of machine learning techniques to overcome the nonlinear nature of this phenomenon.
Deep Learning has provided new avenues of analyzing large amounts of data by applying techniques modeled after biological neurons. These techniques allow for the discovery of nonlinear relationships in massive, complex systems like the data found frequently in fluid dynamics simulation. Using a neural network architecture called Convolutional Neural Networks that specializes in uncovering spatial relationships, a network was trained to estimate the relative intensity of ejection structures within turbulent flow simulation without any a priori knowledge of the underlying flow dynamics. To explore the underlying physics that the trained network might reveal, an interpretation technique called Gradient-based Class Activation Mapping was modified to identify salient regions in the flow field which most influenced the trained network to make an accurate estimation of these organized structures. Using various statistical techniques, these salient regions were found to have a high correlation to ejection structures, and to high positive kinetic energy production, low negative production, and low energy dissipation regions within the flow. Additionally, these techniques present a general framework for identifying nonlinear causal structures in general three-dimensional data in any scientific domain where the underlying physics may be unknown.

Children who are perceived by classmates as being fun increase their peer status over time, but little is known about whether being fun predicts other peer outcomes. Also unknown are mechanisms whereby being fun predicts changes in peer outcomes. Associations with fun, like surgency, suggest that children high in fun are more likely to gain friends while children low in fun are more likely to lose friends, indicating that friend gain and friend loss may serve as intervening links between being fun and peer outcomes. Participants (171 girls, 190 boys) were third to seventh grade students attending a public school in Florida. Across three time points approximately three-months apart, participants reported who their friends were, nominated peers who best fit descriptions of fun and popular, and completed self-reports assessing peer problems. Results indicated that being fun predicted subsequent changes in popularity and peer problems via friend gain but not friend loss. The findings suggest that being fun is a unique predictor of peer outcomes and that friend gain is a mechanism whereby fun children experience positive peer outcomes.

In 1994 when Peter Shor released his namesake algorithm for factoring and solving the discrete logarithm problem he changed cryptography forever. Many of the state-of-the-art cryptosystems for internet and other computerized communications will become obsolete with the advent of quantum computers. Two distinct approaches have grown to avoid the downfall of secure communication: quantum cryptography which is based in physics and information theory, and post-quantum cryptography which uses mathematical foundations believed not to be weak against even quantum assisted adversaries. This thesis is the culmination of several studies involving cryptanalysis of schemes in both the quantum and post-quantum paradigms as well as mathematically founded constructions in the post-quantum regime.
The first two chapters of this thesis on background information are intended for the reader to more fully grasp the later chapters. The third chapter shows an attack and ultimate futility of a variety of related quantum authentication schemes. The fourth chapter shows a parametric improvement over other state-of-the-art schemes in lattice based cryptography by utilizing a different cryptographic primitive. The fifth chapter proposes an attack on specific parameters of a specific lattice-based cryptographic primitive. Finally, chapter six presents a construction for a fully homomorphic encryption scheme adapted to allow for privacy enhanced machine learning.

Access to affordable healthcare is a nationwide concern that impacts most of the United States population. Medicare is a federal government healthcare program that aims to provide affordable health insurance to the elderly population and individuals with select disabilities. Unfortunately, there is a significant amount of fraud, waste, and abuse within the Medicare system that inevitably raises premiums and costs taxpayers billions of dollars each year. Dedicated task forces investigate the most severe fraudulent cases, but with millions of healthcare providers and more than 60 million active Medicare beneficiaries, manual fraud detection efforts are not able to make widespread, meaningful impact. Through the proliferation of electronic health records and continuous breakthroughs in data mining and machine learning, there is a great opportunity to develop and leverage advanced machine learning systems for automating healthcare fraud detection.
This dissertation identifies key challenges associated with predictive modeling for large-scale Medicare fraud detection and presents innovative solutions to address these challenges in order to provide state-of-the-art results on multiple real-world Medicare fraud data sets. Our methodology for curating nine distinct Medicare fraud classification data sets is presented with comprehensive details describing data accumulation, data pre-processing, data aggregation techniques, data enrichment strategies, and improved fraud labeling. Data-level and algorithm-level methods for treating severe class imbalance, including a flexible output thresholding method and a cost-sensitive framework, are evaluated using deep neural network and ensemble learners. Novel encoding techniques and representation learning methods for high-dimensional categorical features are proposed to create expressive representations of provider attributes and billing procedure codes.

Sex Offender Registry and Notification (SORN) and Sex Offender Residency Restrictions (SORR) laws and policies were developed and implemented with the intention of keeping communities and children safe, however many studies illustrate that these policies are in fact resulting in negative consequences for registrants and their families. All the existing studies focused on both registrants and family members, in most cases, spouses. A gap in the literature about the possible impacts on mothers of adult registrants was identified. Further, there was a lack of qualitative research, which, in highlighting the lived experiences and stories of mothers, is powerful and can have a significant impact on increasing social awareness.
The number of registered citizens continues to grow in Florida each year, and there are increasing numbers of families and loved ones of registered citizens that will need supportive and specialized therapeutic services. As a doctoral capstone, this qualitative narrative research involved one-on-one semi-structured interviews with 15 mothers of registered citizens throughout the state of Florida between March and November 2021. The aims of the study were to determine if mothers experienced the same impacts as other family members, or if their experiences were unique to mothering a registered adult child; to learn about how mothers coped with having an adult child on the Florida registry; and finally, what service providers need to know to adequately support this population.

This quantitative study sought to determine the efficacy and mindset perceptions of current school leaders and teachers within public high schools. This study highlighted a discrepancy in efficacy and mindset among educators for each other. Firstly, school leaders feel they make a difference, but teachers do not hold the same level of belief in leadership’s ability to make a difference. School leader perceptions of self-efficacy have increased significantly since 2008. Lastly, teachers’ perceptions of school leader efficacy and teacher mindset correlated, meaning a significant portion of variance in teacher perceptions of school leadership efficacy can be predicted by the mindset held by the teacher toward capacity to grow in ability and talent. However, school leaders’ perceptions of self-efficacy and teacher mindset did not correlate, suggesting school leader self-efficacy beliefs do not predict their beliefs in teacher growth potential. Over the years, the terms “efficacy” and “mindset” have been thoroughly researched; however, never in a context surrounding school leaders’ and teachers’ perceptions of each other’s capabilities. Therefore, this study sought to explore and compare school leaders’ and teachers’ perceptions of efficacy and mindset for each other to gain insight into the workplace environment within educational 9-12 systems.

Autism Spectrum Disorder (ASD) impacts one in every 44 children in the United States (CDC, 2022) and is characterized by marked deficits in social communication with the presence of restricted interests and repetitive behaviors. Students with ASD are increasingly being educated in the general education classroom and are expected to meet the curricular demands thereof (Roberts & Webster, 2020). Due to the core features of their disability, these students often experience significant challenges in written expression. Writing, across content areas, is a primary means in which student learning is measured and evaluated. The COVID-19 pandemic forced educators to explore the use of technology, through the application of synchronous and asynchronous instructional models, to meet the needs of all students while also providing access to Evidence Based Practices (EBPs) and rigorous content (Cox et al., 2021). This study examined the effects of an intervention package consisting of video modeling and virtual coaching on the use of a procedural facilitator (PF) as a planning tool on the overall written quality of the opinion writing with elementary school-aged children with ASD. Writing quality was measured by the presence of planned paragraph elements, Correct Word Sequences (CWS), and Total Words Written (TWW). The significance, acceptability, and effectiveness of the intervention package was also explored.
Results indicate a functional relationship between the intervention package and the presence of planned paragraph elements. The intervention package did not directly impact CWS or TWW. Participant perceptions of the intervention package were generally positive. Caregiver perceptions of the intervention package were generally positive. Implications of the present study are discussed along with limitations and recommendations for future research.